The paradigm of digitalization of agriculture and tillage ...
Digitalization as a Paradigm Changer in Machine-Building ... · Digitalization as a Paradigm...
Transcript of Digitalization as a Paradigm Changer in Machine-Building ... · Digitalization as a Paradigm...
Tampere University of Technology
Digitalization as a Paradigm Changer in Machine-Building Industry
CitationSommarberg, M. (2016). Digitalization as a Paradigm Changer in Machine-Building Industry. (TampereUniversity of Technology. Publication; Vol. 1436). Tampere University of Technology.
Year2016
VersionPublisher's PDF (version of record)
Link to publicationTUTCRIS Portal (http://www.tut.fi/tutcris)
Take down policyIf you believe that this document breaches copyright, please contact [email protected], and we will remove accessto the work immediately and investigate your claim.
Download date:21.05.2020
Matti SommarbergDigitalization as a Paradigm Changer in Machine-BuildingIndustry
Julkaisu 1436 • Publication 1436
Tampere 2016
Tampereen teknillinen yliopisto. Julkaisu 1436 Tampere University of Technology. Publication 1436
Matti Sommarberg Digitalization as a Paradigm Changer in Machine-Building Industry
Thesis for the degree of Doctor of Science in Technology to be presented with due permission for public examination and criticism in Konetalo Building, Auditorium K1702, at Tampere University of Technology, on the 9th of December 2016, at 12 noon.
Tampereen teknillinen yliopisto - Tampere University of Technology Tampere 2016
ISBN 978-952-15-3855-1 (printed) ISBN 978-952-15-3875-9 (PDF) ISSN 1459-2045
1
Abstract
Digitalization is a contemporary societal topic among businessmen, scholars, politi-
cians, and citizens. The way Uber has changed the taxi business and subsequently is
providing new models for the entire transportation industry or even changing urban
planning principles is a practical example of the impact of digitalization. This example
illustrates that digitalization offers major returns for some and ultimate losses for others,
which is similar to Schumpeter’s “Creative Destruction” that he coined in 1942. Digitali-
zation does not refer to a product or service; it is multiple technology-based products,
services, and concepts as a systemic whole. Many of the impacts of digitalization are
difficult to observe beforehand, as the impact rendered is systemic rather than a
straightforward causal relation. Traditional strategic management theories and frame-
works are used to analyze company performance and to explain which strategies indi-
vidual firms or group of firms should implement to succeed. Many of the tools for top
management aid in understanding changes in business environments and offer guid-
ance for making the correct strategic choices, but in many cases, they fail to aid in the
detection of systemic phenomena. At the same time, making these strategic choices is
difficult, as explained by behavioral economics and management cognition, as the
choices involve changing the status quo.
This dissertation examines the digitalization impact on the machine-building industry
that serves global container handling customers - ports and terminals. It is a traditional
capital intensive business-to-business industry that has a relatively small number of
global players. The investigation adopted a value chain view in which machine builders
are actors, actors apply digital technologies provided by enablers. The end customers,
ports and terminals are referred as users. The objective of the research was to in-
crease understanding of digitalization’s potential for disruption or paradigm change as
well as to identify the most important concepts that drive and inhibit this change. As the
change brought about by digitalization is underway, it is necessary to understand
whether the views regarding its impact differ between enablers, actors, and users.
Mixed methods were applied that partly overlapped for triangulation purposes. The
primary methodology included two rounds of Delphi interviews that were complement-
ed by a survey and three case descriptions.
Big Data/Artificial Intelligence emerged as the most prominent digital technology that
can enable disruption in machine-building. Empirical results have shown that Big Da-
ta/Artificial Intelligence challenges the ways knowledge is created; it is more effective
2
when machines and their components are connected to data networks, and the tech-
nology is both rapidly advancing and becoming more affordable. The cost, speed,
availability, and features of Big Data/Artificial Intelligence development are driven by
multiple industries where machine builders can have a relatively small impact.
Empirical results have also shown that discipline and industry-based platforms are the
most powerful economic drivers. The current management of the incumbents has little
experience with these new elements, which have a major influence on industry dynam-
ics. The platforms are especially powerful for change, as they enable a global network
economy in which entrepreneurial knowledge workers can contribute to value creation
in collaboration with startups and multinational corporations. Platform development
cannot be stopped or delayed by incumbents in machine-building. They can ignore the
development, adapt to it, or pursue a platform strategy of their own if the opportunities
match the companies’ capabilities.
Examples of the sub-drivers pushing the digital concepts forward are classical and ra-
tional productivity, lead times, features, quality, and cost. In addition, some of the inhib-
itive sub-drivers are relatively easy to identify, such as 3D printing speed or users
providing access to their data. Concerns regarding data security delay investment, and
changing legacy processes and systems requires time; however, empirical results have
indicated that the strongest inertia is related directly to people and decision making.
Three of the strongest people-related inhibitive sub-drivers are lack of systemic under-
standing, management beliefs, and lack of capabilities. The practical contribution for
management is twofold. First, it must be believed that digitalization will somehow dis-
rupt the current business, and second that the transformation is too complex to be only
planned, but instead requires also experimental learning. A successful combination that
has been suggested by books and articles as well as the results and comments from
the Delphi interviews is developing an entrepreneurial mindset, conducting multiple
small experiments, and applying the knowledge of external networks. This enables
strategy formation through learning, which simultaneously develops the capabilities that
are needed in data and user-centric business environments.
Keywords Digitalization, Industrial Internet, Big Data, Artificial Intelligence, platforms, open
innovation, industry disruption, strategic management, Delphi, container handling
industry
3
Tiivistelmä
Digitalisaatio on ajankohtainen aihe liikemiesten, tutkijoiden, poliitikkojen ja yksittäisten
ihmisten keskuudessa. Yksi käytännön esimerkki digitalisaation vaikutuksista on Uber,
jonka toimintatapa muuttaa taksiliiketoimintaa luoden samalla malleja koko kuljetustoi-
mialalle ja vaikuttaen jopa kaupunkisuunnitteluun. Esimerkki valaisee myös sitä, miten
digitalisaatio tarjoaa merkittäviä voittoja yksille ja kohtalokkaita tappioita toisille, kuten
Schumpeter kuvasi ”luovan tuhon” ajatuksessaan jo vuonna 1942. Digitalisaatio ei ole
yksittäinen tuote tai palvelu, vaan se on tuotteita, palveluita ja konsepteja, joita useat
digitaaliset teknologiat systeemisesti mahdollistavat. Systeemisyys yksinkertaisten
syy–seuraus-suhteiden sijaan tekee vaikutusten ymmärtämisen ja ennustamisen vai-
keaksi. Perinteisiä strategisen johtamisen teorioita ja viitekehyksiä käytetään yritysten
suorituskyvyn analysointiin ja sen ymmärtämiseksi, millä toimenpiteillä yritykset menes-
tyisivät. Lisäksi monet ylimmät johdon työkalut helpottavat näkemään liiketoimintaym-
päristön muutoksia ja tarjoavat tukea oikeiden strategisten valintojen tekemiseen, mut-
ta niissä on heikkouksia systeemisten ilmiöiden havaitsemiseksi. Behavioristinen talo-
ustiede ja johtamisen kognitiotieteet auttavat ymmärtämään, miksi oikeat strategiset
valinnat, jotka muuttavat vallitsevia uskomuksia, ovat vaikeita yksilötasolla.
Tämä väitöstutkimus tutki digitalisaation vaikutusta koneenrakennustoimialaan, joka
palvelee maailmanlaajuista kontinkäsittelyä – satamia ja terminaaleja. Toimiala on pe-
rinteinen, siinä on suhteellisen vähän globaaleja toimijoita ja se sitoo paljon pääomaa.
Tutkimus lähestyi ongelmaa arvoketjun näkökulmasta siten, että koneenrakentajat ovat
toimijoita, jotka soveltavat digitaalisia teknologioita, joita puolestaan mahdollistajat toi-
mittavat. Arvo syntyy lopullisesti käyttäjille, joita ovat satamat ja terminaalit. Tutkimuk-
sen tavoitteena oli lisätä ymmärrystä digitalisaation mahdollisesti aiheuttamasta mur-
roksesta tai muutoksesta nykyiseen arvonluontimalliin sekä siitä, mitkä tekijät hidasta-
vat tätä kehitystä. Koska mahdollinen muutos on meneillään, käsitysten erovaisuuksien
ymmärtäminen arvoketjussa mahdollistajien, toimijoiden ja käyttäjien kesken on tärke-
ää. Tutkimuksen päämenetelmä oli kahden haastattelukierroksen Delfoi-tekniikka sekä
tulosten validiteetin parantamiseksi käytetyt kyselytutkimus sekä kolme case-kuvausta.
Tietomassojen suurtehokäsittely (Big Data) yhdessä tekoälyn (Artificial Intelligence)
kanssa nousi tärkeimmäksi mahdollisen murroksen aikaansaavaksi digitaaliseksi tek-
nologiaksi. Empiiriset tulokset osoittivat, että kyseiset teknologiat vaikuttavat uuden
tietämyksen syntyyn ja että ilmiö kiihtyy, koska koneet ja niiden komponentit liittyvät
kiihtyvässä tahdissa tietoverkkoihin. Kyseiset teknologiat kehittyvät edelleen samalla
4
kun niiden käytön kustannukset laskevat. Nämä teknologiat palvelevat useita toimialoja,
mutta koneenrakentajilla itsellään on vähäinen vaikutus teknologian kustannuksiin,
nopeuteen, saatavuuteen, ominaisuuksiin tai niiden kehittymiseen.
Empiiriset tulokset osoittivat myös, että tieteenhaara tai toimialapohjaiset alustat ovat
voimakkaimmat potentiaalista murrosta aiheuttavat taloudelliset konseptit tämän tutki-
muksen rajauksella. Alustat ovat uusia elementtejä, joilla on merkittävä vaikutus toimi-
alan dynamiikkaan, mutta perinteisten yritysten johdolla on harvoin omaan opiskeluun
tai kokemukseen perustuvaa osaamista niistä. Alustat saavat voimansa verkostovaiku-
tuksista, joissa tietotyöläiset, startupit ja monikansalliset yritykset luovat yhdessä arvoa.
Koneenrakentajat eivät pysty estämään alustojen syntymistä tai merkittävästi hidasta-
maan niiden kehitystä. Ne voivat ohittaa ilmiön, sopeutua siihen tai mahdollisuuksiensa
ja kyvykkyyksiensä puitteissa luoda oman alustastrategiansa.
Merkittävä osa ajureista, jotka kiihdyttävät digitaalisia konsepteja, ovat perinteisiä ja
rationaalisia, kuten tuottavuus, läpimenoajat, ominaisuudet, laatu tai kustannukset. Osa
kehitystä hidastavista ajureista on helposti tunnistettavissa, kuten 3D-tulostimen nope-
us tai kuinka moni käyttäjä antaa pääsyn dataansa. Huoli tietoturvasta hidastaa inves-
tointeja, ja olemassa olevien prosessien ja järjestelmien vaihtaminen on aikaa vievää.
Empiiriset tulokset osoittivat kuitenkin, että voimakkaimmat hidasteet liittyivät suoraan
ihmisiin ja päätöksentekoon. Kolme merkittävintä ihmisiin liittyvää hidastetta olivat sys-
teemisten ilmiöiden huono ymmärtäminen, johdon uskomukset ja kyvykkyyksien puute.
Tulosten merkitys käytännön strategiselle johtamiselle kiteytyy kahteen asiaan. Ensin-
näkin johdon pitää ymmärtää ja uskoa, että digitaalisuus murtaa joiltakin osin nykyisen
liiketoiminnan, ja toiseksi kehitys on niin monisyistä, ettei menestystä voi kovinkaan
tarkasti suunnitella etukäteen. Osa tutkimuksessa käytetystä kirjallisuudesta ja Delfoi-
haastatteluista saadut tulokset painottavat tällaisessa tilanteessa yrittäjyysmäistä ajat-
telutapaa ja paljon pieniä kokeiluja, joissa hyödynnetään ulkoisten verkostojen tietä-
mystä. Toimintatapa mahdollistaa sen, että strategia voidaan luoda oppimalla, mikä
samanaikaisesti kehittää kyvykkyyksiä, joita tarvitaan tieto- ja käyttäjäkeskeisissä liike-
toimintaympäristöissä.
Avainsanat: digitalisaatio, teollinen internet, Big Data, tekoäly, alustat, avoin innovaatio, teolli-nen murros, strateginen johtaminen, Delfoi, kontinkäsittely
5
Acknowledgements
Composing a dissertation at the last quarter of an industrial career is not an obvious
choice. A passion for science, the right topic, support throughout the process, and a
mindset in which beliefs are questioned that have been developed over a lifelong ca-
reer are required. The passion for strategy and the role of people in strategy formation
was inspired by the late Professor Juha Näsi. Juha was a testimony to the vital impact
a great professor has on students’ lives and careers. After applying knowledge in vari-
ous management roles, the inner question arose: what if? The ambiguity and strength
of digitalization also offered unique opportunities in strategic management, including
study with a motivation to learn. The magnitude of the importance of digitalization was
illustrated by the keynote speech of Professor John Zysman from UC Berkeley, which
he gave in Helsinki in the spring of 2012 at the SHOK summit. I thank John for the ini-
tial inspiration but also for his support during my research project. I was also seeking
advice regarding whether an endeavor such as a doctoral dissertation was worth pur-
suing, so I turned to wise men who I knew and trusted. I thank Messieurs Jarl-Thore
Eriksson and Markku Kivikoski, who are known scholars with ample of merits also in
many sectors of the economy and society. You encouraged me and made me renew
the old passion that Juha initially inspired as well as supported me at each meeting
during the journey.
I was still undecided when I met my supervisor Professor Saku Mäkinen. In less than
30 minutes, the decision was clear. Working with Saku was like meeting an old friend
after 10 years and continuing a conversation that was interrupted. The discussions with
Saku have always touched the essence of the research question whether they have
lasted 10 minutes or 10 hours, and I have received excellent guidance when I have lost
track of the theory or methodology. Saku also suggested early on that systemic prob-
lems need systems thinking and introduced me to Professor Matti Vilkko. Matti stimu-
lated my interest in systems thinking methodology and made me regret that I have not
acquainted myself with the methodology before. He has also supported me through
several discussions and provided comments on the systems thinking sub-chapter. I
also wish to express my thanks to Professors Miia Martinsuo, Tommi Mikkonen and
Kari Koskinen from Tampere University of Technology and Assistant Professor Robin
Gustafson from Aalto University. You all have had an interest in my research project
throughout the process and have always been available to discuss both the content
and the methodology of the research.
6
An important issue in this dissertation is open vs. closed systems, whether in business
or in society at large. I was first introduced to this topic at the Finnish Metals and Engi-
neering Competence Cluster (FIMECC), which I was also privileged to serve as Chair-
man of the Board. FIMECC, which has now merged with Digile to form DIMECC (Digi-
tal, Internet, Materials & Engineering Co-Creation), was established by a number of
companies and universities for co-research purposes. While in my role at FIMECC, I
was able to work with both fellow board members and with those responsible for exe-
cuting various duties. I thank them all collectively, but special thanks go to CEO Harri
Kulmala, who has done an outstanding job of leading this community drawn from in-
dustry and academia. Harri has also been a frequent sparring partner in my disserta-
tion work. I also thank Digile (the Need for Speed program) and its former CEO Pauli
Kuosmanen and my own employer, Cargotec, both of which contributed to financing
this research.
Dr. Jari Hämäläinen commented on the digitalization sub-chapter, and Mr. Frank Kho
commented on the sub-section that described the container handling industry—thank
you. Under the supervision of Professor Matti Vilkko, Mr. Tommi Nurmi completed his
master thesis of evaluating statistical methods by using some of the survey data from
this dissertation. After having completed his thesis, Tommi has been keen on discuss-
ing statistical methodology, and he also commented on the subchapter related to the
survey, for which I am thankful. Mrs. Merike Koskinen kindly helped me with the Finn-
ish abstract and my thanks for that. The Delphi panelists provided research input, but in
each interview, the interviewees provided more than asked, and I am thankful to each
one of you. The same way the case descriptions were read and commented by people
who possess deep knowledge from the case firms. The names of the Delphi panelists
are listed in the exhibit and case contributors in the respective case descriptions.
I also wish to thank pre-examiners Professor Emeritus Alok Chakrabarti from the New
Jersey Institute of Technology and Adjunct Professor Osmo Kuusi from Aalto University
and Turku University. Even when I thought that the manuscript was ready, they found
new perspectives that I had overlooked, and I thank they for these valuable contribu-
tions.
You can blindly follow the rules, or you can think; I acquired that mindset from my first
superiors in the workplace, who also turned out to be my mentors. I thank the late Mr.
Terence Derry for what I learned about people relations in a business context. That
was during the period when What They Don’t Teach You in Harvard Business School
was published, and I learned its lessons from him. Questioning existing paradigms re-
7
quires a culture in which there are no bad questions. Under the leadership of Mr. Rai-
mo Ylivakeri, I learned a leadership style that encouraged innovation. Last but not least,
I would like to thank Mr. Christer Granskog, who not only strengthened the foundation
of my knowledge but also provided an example of outstanding general management.
He did not hide behind complexity or uncertainty; he mastered long-term and short-
term goals and issues while exemplifying the role of management when faced with con-
flicting expectations among stakeholders. Most importantly, I learned how to utilize
strategic thinking in day-to-day management, inspired initially by Professor Juha Näsi.
Over a 30-year career, a network of colleagues and business partners accumulate.
Many of them have been available for conversation on issues during this research or
otherwise expressed support during my moments of frustration, and I offer my collec-
tive gratitude to all of them.
Life is full even without a dissertation; pursuing this type of a dream might be a bit self-
ish. I have at times been short-fused, mentally absent, and likely have fallen short in
ways that I do not even realize. Thus, I thank my wife Riitta and sons Marko and Mika
for their love, support, and patience. They were all a critical source of support regarding
some of the thoughts that I have shared during the journey.
My final thanks go to my mother Kaisu and father Aimo, who have always trusted me to
do things my way.
.
8
Table of Contents
Abstract
Tiivistelmä
Acknowledgements
List of Figures
List of Tables
List of Abbreviations
List of Key Concepts
1 INTRODUCTION ............................................................................................... 19
Motivation...................................................................................................... 20 1.1
Research questions ...................................................................................... 23 1.2
Structure of the dissertation .......................................................................... 25 1.3
2 THEORETICAL FOUNDATIONS ....................................................................... 27
Industry evolution .......................................................................................... 28 2.1
2.1.1 Industry definition ................................................................................. 28
2.1.2 Industry evolution ................................................................................. 31
2.1.3 Disruptive vs. sustaining change .......................................................... 36
Economy, strategy formation and top management teams ............................ 41 2.2
2.2.1 Knowledge based economy ................................................................. 41
2.2.2 Strategy formation ................................................................................ 44
9
2.2.3 Top management team ........................................................................ 49
2.2.4 Value chain and process thinking ......................................................... 57
2.2.5 Industry beliefs ..................................................................................... 60
2.2.6 Business models .................................................................................. 61
Digitalization and strategy ............................................................................. 63 2.3
2.3.1 Platform strategy .................................................................................. 65
2.3.2 Open innovation as a paradigm in the development of a firm ............... 66
2.3.3 The enabling technology behind digitalization ...................................... 69
2.3.4 Digitalization as a systemic phenomenon ............................................. 72
Synthesis on industry evolution, strategy formation, and digitalization ........... 81 2.4
3 METHODOLOGICAL FOUNDATION ................................................................. 84
Research philosophy ..................................................................................... 84 3.1
Mixed methods .............................................................................................. 87 3.2
Delphi method ............................................................................................... 88 3.3
3.3.1 Delphi round one .................................................................................. 92
3.3.2 Delphi round two .................................................................................. 94
Survey ........................................................................................................... 96 3.4
3.4.1 Experiment ........................................................................................... 96
3.4.2 Analysis ............................................................................................... 98
Cases .......................................................................................................... 100 3.5
4 RESULTS ........................................................................................................ 102
Global container handling industry .............................................................. 102 4.1
10
Results from the Delphi interviews .............................................................. 108 4.2
4.2.1 First Delphi interview round ................................................................ 109
4.2.2 Second Delhi interview round ............................................................. 116
4.2.2.1 Technological drivers ..................................................................... 116
4.2.2.2 Economic / strategic management drivers ..................................... 127
Results from the survey ............................................................................... 138 4.3
4.3.1 Whole population ............................................................................... 140
4.3.2 By the value chain .............................................................................. 142
4.3.3 Dependencies .................................................................................... 144
Cases .......................................................................................................... 146 4.4
4.4.1 Enevo................................................................................................. 147
4.4.2 Trimble ............................................................................................... 149
4.4.3 XVELA ............................................................................................... 151
4.4.4 Summary of the three case descriptions ............................................ 154
5 DISCUSSION AND CONCLUSIONS ............................................................... 155
Discussion ................................................................................................... 156 5.1
5.1.1 Disruption ........................................................................................... 156
5.1.2 Drivers ............................................................................................... 162
5.1.3 Value chain ........................................................................................ 169
5.1.4 The systemic impact .......................................................................... 172
Conclusions ................................................................................................ 174 5.2
5.2.1 Industry evolution, strategy formation ................................................. 175
11
5.2.2 Practitioners in machine-building industry .......................................... 177
Validity and reliability ................................................................................... 180 5.3
Future research avenues ............................................................................ 187 5.4
REFERENCES ......................................................................................................... 189
APPENDIXES
Appendix 1. Mapping strategic schools related to strategy formation (Minzberg et al.
1998) ........................................................................................................................ 212
Appendix 2. Delphi round two interview framework ................................................... 213
Appendix 3. Survey form ........................................................................................... 221
Appendix 4. Participants in the Delphi round one and/or two .................................... 222
Appendix 5. Results from Shapiro-Wilk test .............................................................. 224
Appendix 6. Interaction effect test ............................................................................. 225
Appendix 7. Disruption impact by whole population, unbalanced data ...................... 226
Appendix 8. Disruption impact by value chain, balanced data ................................... 227
Appendix 9.Disruptiveness of individual technologies or concepts by value chain,
balanced data ........................................................................................................... 228
Appendix 10. Correlation matrix (Kendall) with whole population .............................. 230
Appendix 11. Correlation matrix (Kendall) by value chain with balanced data ........... 231
Appendix 12. Variance dependency by ANOVA ........................................................ 232
12
List of Figures
Figure 1. The research process .................................................................................. 26
Figure 2. Network resources, mechanism, and expected relationships (Lee, 2007) .... 35
Figure 3. Impact of sustaining and disruptive technological change (Christensen, 2011)
................................................................................................................................... 37
Figure 4. Multitier visual output of a technology roadmap (Walsh, 2004) ..................... 40
Figure 5. Outcome Economy concept (World Economic Forum, 2015) ....................... 42
Figure 6. Typology of innovation by strategic focus (Deschamps, 2005) ..................... 55
Figure 7. Value chain (Porter, 1985) ........................................................................... 57
Figure 8. Sources of Dominant Logic (Prahalad and Bettis, 1986) .............................. 61
Figure 9. Business Model Canvas (Osterwalder and Pigneur, 2010) .......................... 62
Figure 10. Maturity by industry, (Westerman et al., 2012) ........................................... 81
Figure 11. The research object ................................................................................... 83
Figure 12. The research design .................................................................................. 88
Figure 13. Ship size development in TEU's (Drewry Maritime Research, 2014) ........ 107
Figure 14. Preliminary finding on systemic characteristics of disruption .................... 114
Figure 15. Three highest ranked technological drivers in the Delphi panel in round two
................................................................................................................................. 117
Figure 16. Three highest ranked technological drivers in the Delphi panel in round two
by the value chain ..................................................................................................... 118
Figure 17. Top eight aggregate technological sub-drivers ......................................... 125
Figure 18. Summary of the technology drivers by the value chain............................. 126
13
Figure 19. Three highest ranked economic / strategic management drivers in the Delphi
panel in round two .................................................................................................... 128
Figure 20. Three highest ranked economic / strategic management drivers in the Delphi
panel in round two by the value chain ....................................................................... 129
Figure 21. Top eight aggregate economic / strategic management sub-drivers ......... 136
Figure 22. Summary of the economic / strategic management drivers by the value
chain ......................................................................................................................... 137
Figure 23. Timing of potential disruption by whole population, unbalanced data ....... 141
Figure 24, Weak and strong disruption impacts by the value chain, balanced data ... 142
Figure 25. Weak and strong disruption drivers by value chain, balanced data .......... 143
Figure 26. Impact of own digitalization investments, balanced data .......................... 144
Figure 27. Big Data /AI and Networks, Crowds, Platforms described by CLD ........... 173
14
List of Tables
Table 1. Four worldviews (Creswell, 2003) ................................................................ 86
Table 2. Three types of experts on future developments (Kuusi, 1999) ....................... 91
Table 3. Five largest shipping lines (Alphaliner, 2016) .............................................. 104
Table 4. Five largest terminal operators (Drewry Maritime Research, 2014) ............. 104
Table 5. The Delphi panel in round one .................................................................... 109
Table 6. Results from Delphi round one, technological drivers .................................. 111
Table 7. Results from Delphi round one, economic / strategic management drivers . 113
Table 8. Rank order of the technological drivers in Delphi round two ........................ 116
Table 9. New technological sub-drivers from the Delphi round two ........................... 119
Table 10. Rank order of the economic / strategic management drivers in Delphi round
two ............................................................................................................................ 127
Table 11. New economic / strategic management sub-drivers from the Delphi round
two ............................................................................................................................ 130
Table 12. Participation in the survey ......................................................................... 138
Table 13. ANOVA F-test results ................................................................................ 145
Table 14. Summary of the three case descriptions ................................................... 154
15
List of Abbreviations
ADC Analog Digital Converter
AI Artificial Intelligence
AM Additive Manufacturing
B2B Business to Business
B2C Business to Consumers
CAD Computer Aided Design
CLD Causal Loop Diagram
CPS Cyber-Physical Systems
CRM Customer Relation Management
EBIT Earnings Before Interest and Taxes
EDI Electronic Data Interchange
ERP Enterprise Resource Planning
GDP Gross Domestic Product
GPS Global Position System
GSM Global System for Mobile communication
HCI Human to Computer Interface
ICT Information and Communication Technology
IoT Internet of Things
IP Internet Protocol
IPR Intellectual Property Rights
JV Joint Venture
LCC Life Cycle Costs
NAISC North American Industry Classification System
MBSE Model Based System Engineering
16
MEMS Microelectromechanical systems
MNE Multi National Enterprises
M2M Machine to Machine communication
OS Operating System
R&D Research and Development
SIC Standard Industrial Classification
SME Small and Medium Sized Enterprises
TEU Twenty-foot Equivalent Unit
VAS Visual Analogue Scale
VOIP Voice over Internet Protocol
W-LAN Wireless Local Area Network
Note All words of explicit concepts or theories begin with capital letters, such
as Big Data or Bounded Rationality
17
List of Key Concepts
Actor A machine-builder, who integrates either the technologies of
external suppliers or their company’s own technologies in order to build machines using several engineering disciplines.
Continuous improvement Incremental improvement as a result of a development activ-
ity. It is assumed that in competition, continuous improve-ment will ensure that a firm maintains its current position rel-ative to its peers.
Digitalization The use of digital technologies to create value for a firm.
Digitization (as opposed to digitalization) can also be used in this way, but here this term is strictly considered as relating to the technological process of changing analogical data into a digital form.
Disruption The paradigm change in rules concerning how value is cre-
ated in business. Driver A factor that enables potential disruption in the machine-
building industry. Drivers can be either digital technologies or economic/ strategic management concepts that utilize op-portunities that the digital technologies enable.
Enabler A firm that supplies digital technologies, which can be prod-
ucts or services, to actors and users. Paradigm The entire constellation of beliefs, values and techniques,
and so on shared by the members of a given community (Kuhn, 1996). In this context paradigm relates to paradigm of value creation of a firm.
Sub-driver A factor that inhibits or accelerates the disruptive impact of
the driver. Quantum leap A major change in the market position as a result of a de-
velopment activity. Typically, a quantum leap is the result of an innovation.
User A firm, whose business is to handle containers at ports and
terminals.
18
“Be the change that you wish to see in the world.”
Mahatma Gandhi
19
1 Introduction
Communities have sought ways to achieve objectives that they perceived as valuable
for thousands of years, such as conquering new regions, cultivating land, or in modern
times, developing and delivering products and services in exchange for financial re-
wards to be distributed to stakeholders. This journey has always included fixed charac-
teristics, such as people and their relations, hierarchies, culture and social codes, other
sets of people with the same goals, and state-of-the-art tools and methods.
The discipline that is dealing with the way of achieving these objectives is usually
called strategy both in academic world and in common language. Sun Tzu expressed
his thoughts about strategy in the Art of War over 2000 years ago (Sun, 1963). Leader-
ship as a means to achieve goals is not specifically a new invention, either; Niccolò
Machiavelli’s Il Principe was published in 1532 (Machiavelli, 2003), and like Sun Tzu’s
book, it is read by business leaders in the current millennium. The Wealth of Nations
published in 1776 by Adam Smith was a milestone for modern economics (Smith, 2014)
in the same way that The Principles of Scientific Management by Fredrick Taylor in
1911 was for modern business administration (Taylor, 1998). Innovation has always
produced unexpected results, causing disruptions or paradigm changes. The list of
disruptive innovations is endless: fire making, bows and arrows, steam engine, spin-
ning jenny, telegraph, transistor, Internet, etc.
Likewise, learning by experimenting and sharing knowledge is as old as mankind
(Wenger and Snyder, 2000). The major disruptions in knowledge creation are the in-
vention of writing by the Sumerians in 3000 BC (Fischer, 2008) and the invention of
printing by Johannes Gutenberg in 1448 (Childress, 2008). In this dissertation, it is as-
sumed that the ability to transform any type of data, observations, events, etc. to digital
20
forms as well as the ability to process it, store it, and share it economically, globally,
and in real time have the same type of systemic characteristics as the invention of writ-
ing or printing.
Is digitalization as disruptive as Uber in taxi business suggests? Do the strategic man-
agement schools and tools detect the systemic change in business environment in way
that that can be used for strategy formation? The objective of this dissertation is to in-
crease understanding on the potential impact on digitalization on machine-building in-
dustry and factors that either drive or inhibit the development.
Motivation 1.1
The motivation for this research included also personal inspiration. I have been fortu-
nate to witness the transformation of a Finnish state-owned aircraft factory to a tech-
nology center that is a global leader in the development of automated ports and termi-
nals over the last 30 years. I have been part of the transformation in senior manage-
ment by leading functional or business organizations and part of the corporate staff
responsible for strategic planning, mergers and acquisitions, and technology. This jour-
ney brought me either temporarily or for longer periods to one-third of the world’s coun-
tries, which has increased my understanding of globalization on the macro level and of
people with different backgrounds on the micro level.
The factory celebrated its 75th year anniversary in 2011, and Cargotec published the
history of the factory (Koivuniemi, 2013), which was also a story of survival during sev-
eral disruptions beginning with war and war reparations. The pace of change acceler-
ated in the 1980s. The product offerings narrowed from tens of categories to half a
dozen, and the factory eventually began to serve only one customer industry: ports and
terminals that handle containers. The transformation also involved a geographical
change from a domestic supplier to a center that is a part of a global network. The in-
ternationalization generally followed the pattern described by Luostarinen (1979) in
which products, operations, and markets have their own internationalization patterns
but jointly contribute to the development, growth, and mature phases of the internation-
alization of a firm. At the same time, the value chain narrowed from a fully integrated
factory to a competence center (research and development, sales and marketing, pro-
ject management, services, and support functions). Between 1994 and 2005, the facto-
ry underwent five ownership changes, including several changes in the company and
21
product names. The last milestone was in 2012 when the factory moved to a purpose-
built technology center, including a large test site, laboratories, and a workshop for pro-
totypes, which is located near Tampere University of Technology in Finland.
The common denominator of survival was versatile technological competence through-
out the organization. The core competence related to machine-building, and the digital
technology adaptation was evolutionary. Information technology came first to adminis-
trative processes, and then it was embedded in production machinery. In the 1980s,
programmable logic was introduced in the products. Today’s autonomous machines
represent a continuum of that development. They are highly sophisticated, but the fun-
damentals of business have remained the same in comparison with digitalization in the
media business. A popular claim is that digitalization has a marginal impact on capital
intensive industries. Increasing the understanding of this claim was the personal inspi-
ration behind this dissertation, which is also relevant in determining whether there is a
need for the next transformation of that former aircraft factory.
The Internet of Things, Industrial Internet, and Big Data are all examples within the
digitalization domain. They have all appeared in recent years in the Gartner’s hype
curve (LeHong et al., 2014). McKinsey Global Institute (Manyika et al., 2013) published
a research report in which 12 disruptive technologies in order of their estimated poten-
tial economic impact by year 2025 were listed. The first six are relevant to this disserta-
tion. General Electric (GE), which is considered the father of the Industrial Internet con-
cept, described the fundamental logic for a machine builder in their white paper (Evans
and Annunziata, 2012). In table 1 on page four, and example is used in which it is es-
timated that the value of 1 % fuel savings in commercial aviation would amount to $30
billion in savings over 15 years. This could be achieved using the Industrial Internet,
and the benefit could be shared between the developers and the users.
These data suggest that digitalization would be highly relevant for all machine-builders,
but it does not necessarily indicate industrial disruption. For many companies, the main
digitalization agenda is related to the growth opportunity that it offers; however, there
are ample disruption examples from other industries, such as the management beliefs
of Kodak in films vs. digital photography (Lucas and Goh, 2009), IBM as an example of
surviving disruption through transformation (Gerstner, 2002), and Amazon disrupting
bookstore businesses with digitally enabled customer insight and superior logistics
(Kimble and Bourdon, 2013). Every industry has its special characteristics and it also
easy to state these examples are not applicable to one’s own industry. To better un-
derstand the discussions that took place in management in the industries that have
22
experienced digital disruption, senior executives from three industries who were insid-
ers in the digitalization discussion before the disruption took place in their respective
industries (media, fine paper, and retail) were interviewed.
An Editor-In-Chief of a large newspaper explained the media’s experience with digitali-
zation. In the early 1990s, newspapers were “ready”: printing was effective, distribution
was in order, readers were satisfied with the content, and owners received good re-
turns on their investments. At the same time, local language was protecting national
newspapers from international competition, and the domestic market was stable be-
tween the national leaders and regional newspapers. The production process was the
first to be digitalized. Then digitalization emerged in the form of internet, emails, mobile
phones, and laptops, but they were tools used to complete the same processes as be-
fore. The new millennium experienced a change driven by owners’ ambitions and the
dot.com era. Several visionary experiments were performed, but they were buried at
the same time as dot.com had its crash landing. Google and YouTube were identified
as potential future competitors but were underestimated, partly due to several parallel
changes at the time and partly due to earlier failed experiments that might have been
premature. The disruption caused by digitalization also occurred during a low cycle, so
the focus was on cutting costs. There was also considerable natural inertia, as the ma-
jority of the assets and personnel were in printing and distribution. Incumbents were not
able to translate the experiments to the next level, which was eventually done by play-
ers outside of the industry, such as Google and Facebook; however, this has also
pushed the surviving incumbents toward digitalization maturity that surpasses most
other industries.
Electronic mail did not create a “paperless office,” a slogan from more than 15 years
ago, which could be the reason paper companies did not experience the magnitude of
digitalization in their industry. A retired executive from a large Finnish wood processing
company stated that the impact on paper demand was underestimated simply because
improvements in old strategies still paid off, which in practice referred to investments in
faster machines or shutting older machines down. The weak signals of digitalization
existed but they were more apparent to colleagues whose business was not dependent
on the fine paper. In hindsight, it is obvious that investments in new raw material-based
products would have been a good option. This scenario was available, but it was shad-
owed by the old paradigm. The focus has now shifted, but the lead times for success
are lengthy.
23
The retail example is analogical. Currently, shops and department stores are in the
midst of closing; however, many of these stores were built when e-commerce was al-
ready changing consumer habits. A former retail executive admitted that this situation
was foreseeable over 10 years ago. The impact of digital development was acknowl-
edged, and there were several experiments using digital technologies that had been
conducted; however, as previously mentioned, the magnitude of the impact was under-
estimated. The root cause for the underestimation was the same as in media: digitali-
zation was predominantly operating under the old paradigm. In retail, this meant that
Big Data was utilized to optimize the supply chain or that information was used for bar-
gaining with suppliers. To a lesser degree, it was used for improving marketing com-
munication with customers, recognizing changing consumer behaviors, and innovating
business models.
Practitioners have acknowledged the potential of digitalization. A growing number of
digital strategies have been implemented in firms, and experiments have been initiated.
There is a justified assumption that digitalization also has an impact on firms’ compe-
tencies. Old competencies change or disappear by automation, which is both a mana-
gerial and a societal issue. At the same time, the need for new types of competencies
has emerged. A study by Oxford University (Frey and Osborne, 2013) has suggested
that the next wave of computerization will impact also knowledge work. Also new ways
of working emerge in startup companies. This is one reason that even large multina-
tional companies have begun to use learnings from startups, which is evident in GE’s
CEO Jeffrey Immelt’s explanation for hiring 1,000 people from Silicon Valley:
“If you give them the room and pick the right idea, you can hire great people with a great mission and you can change a company.”1
Research questions 1.2
This dissertation is based on an assumption that digitalization will change some of the
fundamentals regarding the way that machine-builders create value during the next 10
years. The potential changes in paradigms are assumed to be systemic, but to obtain
theoretical and empiric evidence, the research question was divided into sections. The
1 GE CEO Jeffrey Immelt in The World Energy Innovation Forum, Pando Daily 15.5.2014
24
objects of the impact are categorized as products, services, operations, and business
models. The factors that enable the potential impact are called drivers, and they can be
digital technologies or economic/ strategic management concepts that utilize opportuni-
ties that digital technologies enable. The inquiry focused on disruptive impacts, which
can be referred to as paradigm changes or as changing the rules regarding how value
is created in an industry. It is an important distinction from quantum leaps in market
positions, which is a normal objective of major R&D projects of the firms. For most
industries, the largest share of development money is invested in continuous improve-
ment because in tough competition keeping the current market position requires al-
ready a lot of effort.
Machine-builders are called “actors” that use the technology of several domains and
that use engineering disciplines to integrate those technologies to solve the needs of its
customers, which in this dissertation are called “users.” Firms that supply digital tech-
nologies that can be products or services are called “enablers.” Actors might have their
own core technology and might make multiple choices regarding how they compete
within the market they have selected. A typical choice for actors involves which part of
the value chain it opts to have own resources. To narrow down the empirical portion of
this dissertation, the actor must have following characteristics:
The solution of the actor is a long-term investment for the users regardless of
whether it is accounted for on the balance sheet or considered a cost item in
the profit and loss statement
The actor operates in a market in which there is a global supply and demand
The solution of the actor in the user process is critical and consequently up-
time is critical
The machines have already autonomous versions or those will likely be avail-
able within the next 10 years
The last “in scope” criteria is particularly important, as the autonomous vehicles utilize
various types of digital technologies. Autonomous operation is also quite different for
users, as the important part of the users’ value has been dependent on the actions of
an operator (often a driver). It is also a change to the development paradigm of the
machines for the actor. An important part of machine behaviour data has been based
on drivers’ experiences. Autonomous machines create data that is measurable. It is
often real-time data, and it can be perceived as factual. This type of data is a new
source of input that the actor can use when developing new products, services, opera-
tions, and business models. Autonomous operations have though existed for decades
in the process industry.
25
Finally, the value is captured by the user. In an industrial context, the user can be a
power station, a chemical plant, or a service provider. The empirical portion of this dis-
sertation assumed that users are global container handling operators at ports. A port
operator uses systems, machines, and services to serve its customers, which include
shipping lines at sea and trucking and railroad companies on land.
The primary assumption is:
Digitalization is a major systemic enabler for future competitive ad-
vantages of the machine-building industry, and in some phase of the val-
ue creation, it is or will be disruptive within the next 10 years.
The research questions are:
Which drivers enabled by digitalization accelerate or inhibit disruption in a
machine-building industry that serves global container handling?
Do the enablers, actors and users have different views on the drivers or
sub-drivers that either accelerate or inhibit the disruptive development?
How are disruptive drivers related to each other in a machine-building in-
dustry that serves global container handling?
The answers to these questions will provide support for decision-makers regarding
their current strategic choices and allocation of resources. Ultimately, the benefit mate-
rializes in better value creation or in extreme disruption with a higher likelihood to stay
in business or to lead the change.
Structure of the dissertation 1.3
Chapter two describes the theoretical foundation for the empirical research. The re-
search question is considered from three perspectives. Industry evolution, disruption,
and convergence address the question from an external viewpoint. Firms take actions
in businesses, and the strategic decisions are usually made by management. There-
fore, the second perspective of the theories is derived from economics and strategic
management. The characteristics of digitalization might explain some of the disruption
potential. Digitalization has also triggered new strategic concepts, such as platforms,
26
and thus the third perspective of the theories involves digitalization. The inquiry fo-
cused on systemic changes, which resulted in a rather wide theoretical range
The third chapter presents the research methodology. Mixed methods were utilized,
and the primary method used is the Delphi technique, which was triangulated by a sur-
vey and three case descriptions. Findings from thematic Delphi round one served as
input for Delphi round two. They were also used to fine-tune the survey and case de-
sign. The survey was also used to expand the coverage of the Delphi interviews.
The results are presented in chapter four so that each method forms an individual sec-
tion. The chapter begins with a short description of the container handling industry to
provide a background for the interpretation of the results. The discussion and conclu-
sions combine qualitative (Delphi interviews and cases) and quantitative (survey) re-
sults in chapter five. Also the validity and reliability is reviewed in this chapter. The
chapter concludes by summarizing key contributions and suggestions for future re-
search avenues. The structure of the dissertation is described as a process in Figure 1.
Figure 1. The research process
Objective
Specifica-tion
Result
Industry evolu-tion Strategy and top management Relevant digital-ization concepts
Framework for the empirical research
Theoretical foundation
Design meth-odology based on the inquiry
Mixed methods
* Delphi * Survey * Cases
Framework and methods for the empirical re-search
Validate drivers Suggest accelera-tive and inhibitive sub-drivers
Analyze the po-tential disruption, its drivers and sub-drivers
Discuss, conclude and validate results Suggest new re-search avenues
Derive potential drivers of dis-ruption based on theory
Methodological
foundation Delphi 1
Answers to open questions based on the theoreti-cal framework
Industry refined framework for the Delphi round 2. Survey and case design
Discussion Conclusions
* Disruption impact * Critical drivers * Critical sub-drivers * System model * Contribution to theory and practice * Future research avenues
Disruption analysis by the value chain
Leading expert judgment * Delphi panel * Survey partici-pants Case descriptions
Triangulation between the three methods, theory, and in-dustry knowledge
Delphi 2, Sur-vey, Cases
27
2 Theoretical foundations
The inquiry involved investigating potential industrial disruption, its drivers, and its in-
hibitors. A general assumption regarding disruption in business is that somehow the
rules regarding how value is created change. More generically, it is similar to a para-
digm change.
Thomas Kuhn (1996) defined the usage of a paradigm:
On the other hand, ‘paradigm’ stands for the entire constellation of beliefs, values and techniques, and so on shared by the members of given community. On the other hand, it denotes one sort of element in that constellation, the concrete puz-zle-solutions which, employed as models or examples, can replace explicit rules as a basis for the remaining puzzles of normal science.
A new innovation often acts as a catalyst in this type of a change. In this dissertation,
the catalysts that were studied were related to digital technologies. Bower and Chris-
tensen (1995) characterized technologies that can be disruptive by stating:
The technological changes that damage established companies are not usually radically new or difficult from a technological point of view. They do, however, have two important characteristics: First, they typically present a different pack-age of performance attributes—ones that, at least at the outset, are not valued by existing customers. Second, the performance attributes that existing customers do value improve at such rapid rate that the new technology can later invade those established markets.
The first set of theories focuses on how industries evolve. In that evolution, changes
initiated by businesses largely depend on decisions made by a firm’s management.
Therefore, the second set of theories involved both strategic management and theories
that focus on managers as part of the organizations of those firms. To understand the
28
potential of digitalization as well as the drivers and inhibitors, the chosen digital tech-
nologies were reviewed from a systemic perspective. There is an assumption that digi-
tal technologies offer a wide range of systemic concepts that impact both the external
business environment and the internal processes and capabilities of firms. This led to a
situation in which the three main theory areas consisted of a large number of individual
perspectives. To keep the perspectives manageable, there was an attempt to focus on
the seminal theories of each perspective; the judgement on selection was either based
on academic citations or research indicating major usage by the practitioners. Chapter
two concludes by merging the three sets of reviewed theory areas with the framework.
Industry evolution 2.1
This dissertation focuses on a machine-building industry that serves global ports and
terminal customers, and therefore this sub-chapter begins with the perspectives of es-
tablished industries. Thereafter, theories that explain how industries evolve are re-
viewed. Evolution can be either incremental or radical. Particular interest is paid to the
players, components, and mechanisms that are involved in industry dynamics.
2.1.1 Industry definition
The broad definition of an industry, “a particular form of branch of economic or com-
mercial activity,”2 can lead to interpretations based on the type of activity, such as a
service industry, or the product, such as the automobile industry. Established industries
are already categorized. The most practical categories can be found by viewing lists of
classified industries from federations or by examining companies that are grouped un-
der different industries based on the stock exchange. The United States has a pivotal
position in trading stocks or other financial instruments of companies. The U.S. Securi-
ties and Exchange Commission is the governing authority, and it uses SIC codes
(Standard Industrial Classification) when categorizing companies by industries.3 SIC
coding is relatively fine-tuned; however, the category to which the focus of this disserta-
2 http://www.oxforddictionaries.com/definition/english/industry, retrieved 6.6.2016
3 http://www.sec.gov/info/edgar/siccodes.htm, retrieved 6.6.2016
29
tion belongs (3510 = construction, mining & materials handling machinery & equipment)
is not exhaustive when considering the needs of users. One way to solve this issue is
moving towards more aggregate grouping and using the NAISC (North American In-
dustrial Classifications System) in which the industry would be classified into the group
of 31-334, which is manufacturing. This would include the needs of the user but would
also include several products that are not relevant. Despite the weaknesses of both the
SIC and the NAISC, similar categories offer suitable platforms for research within the
industry or between industries, especially if statistical methods or time series are ap-
plied. The same longevity that simplifies the trend analysis might create inertia that
delays the renewal. This could be explained by the multiple stakeholders that have
emerged based on classifications, such as financial analysts, standardization organiza-
tions, trade associations, employee organizations, labor unions, conferences, and trade
magazines. The convergence of industries blurs the boundaries between old industries
and emerging new industries.
Some components appear to be characteristics in the theory of industry. Components
such as supply, demand, price, and competition are rooted in the very principles of
economic theories, even if the details differ depending on which economical system
they represent. The division of labor (Smith, 2014) was originally based not only on
specialization but also on the high amount of the value being in a manual work in the
industrial production. The evolution of services, and later, knowledge-based services,
has brought several intangible assets to complement the original factors of production:
labor, raw materials, and capital. Products and services fulfill needs in the market, but
they do have substitutes. Substitutes can be different products, methods, or services
that fulfill the same need, which often is the explicit object of innovation. Differentiation
is a smaller degree of substitution that is also important in avoiding price competition
(Conner, 1991). One of the main conclusions in her article (Conner, 1991), which com-
pared five schools of industrial organization economics, was the importance of unique
capabilities, which are quite different when considering labor to be a homogenous fac-
tor of production. The impact of a non-economic environment on economic develop-
ment has been discussed since the infancy of the economics field. Schumpeter (2012)
highlighted also that it is wrong to simplify economic development by decoupling it from
the surrounding societal development.
4 http://www.census.gov/cgi-bin/sssd/naics/naicsrch, retrieved 6.6.2016
30
One of the most often cited modern theories used to explain industry dynamics is Mi-
chael Porter’s Five Forces (1979). The supply (bargaining power of suppliers) and de-
mand (bargaining power of customers) form the core of the model. Substitutes in eco-
nomics have been translated as a threat of new entrants or a threat of new products
and services. These forces impact the dynamics that already exist between the current
competitors of a particular industry.
The Schumpeter’s (2012) societal view of in economic development was not included
in the Five Forces. Thomas Freeman’s (2015) stakeholder model paved the way for
today’s sustainability as an explicit form of business practice. The list of stakeholders is
extensive: customers, competitors, media, employees, special interest groups, envi-
ronmentalists, suppliers, governments, local community organizations, owners, and
consumer advocates. As an extension to Porter’s (1979) Five Forces, Freeman (2015)
suggested that adding stakeholders as a sixth force would complement the model.
Laplume et al. (2008) conducted an extensive literature research based on 179 articles
between 1984 and 2007 and concluded that past stakeholder research was divided into
five categories: Identification and Salience, Stakeholder Actions/Response, Firm Ac-
tions/Response, Firm Performance, and Theory Debates. The Identification and Sali-
ence group was closest to the industry definition, and there were 19 theoretical and 13
empirical articles during the period examined. Mitchell et al. (1997) studied the typology
and salience of the stakeholder theory for the management of a firm and determined
that stakeholders must possess power, legitimacy, and urgency, and the strength of the
impact depends on the number of attributes that are simultaneously present. In both
papers, the view of stakeholders is firm-centric, and they do not capture the stakehold-
er role as a source of potential industrial disruption. The same pattern is evident in arti-
cles at large; searching “Stakeholder” in Web of Science (6.6.2016) provides 129290
results, “Stakeholder, Industry” provides 14795 results, and “Stakeholder, Industry,
Disruption” only provides 43 results.
The Schumpeterian (2012) view was also present in the PEST (Political, Economic,
Social, and Technological) model (Aguilar, 1967). It is firm-centric and was not de-
signed as a tool to diagnose industrial disruption, but it describes the factors that ac-
cording to Aguilar (1967), form sources of external strategic information that top man-
agement must understand for strategic planning. The PEST model has also got deriva-
tives when scholars have added e.g. legal, environmental, or ethical factors to the orig-
inal model. The limitations of PEST and its derivatives were discussed by Burt et al.
(2006) in the context of scenario planning. The paper emphasized that the business
31
environment is also an outcome of management thinking. In addition, it suggested that
the their scenario methodology contains a process of sense making that can identify
discontinuities in the environment better than applying the static PEST model (Burt et
al., 2006).
2.1.2 Industry evolution
Emergence suggests that something new is being created; evolution is normally ac-
companied by development, which is a continuum of something that already exists.
Disruption refers to discontinuity in development, whereas convergence is a special
case, as the mechanism of disruption relates to a merger of two or more existing phe-
nomena. Innovations based on digital technologies can most likely be a catalyst in all of
the four above mentioned forms of industrial evolution.
The question of emergence is existential; what was there before the industry was born?
In the same way, scientific discipline industries can be formed by allowing a subset of
existing industry to form its own industry. Logically, one of the contributors is new
knowledge, or innovation. If a new medicine is invented, the extent of the application
required to form a new subindustry from more generic pharmaceuticals is a matter of
definition. It is unlikely that a firm’s innovation activity is targeted to form new industries;
rather, the aim is most likely to satisfy the new or existing needs of the market. Aber-
nathy and Utterback (1978) discussed the importance of making a distinction between
evolutionary and radical innovation. The distinction cannot exclude the ex-ante incre-
mental activity that yields an unplanned radical impact as a side effect.
Although the Five Forces (Porter, 1979) can explain industry disruption, the model was
developed at a time when the physical value chain held more importance. Porter (1979)
also highlighted strategy formation based on the understanding of the industry evolu-
tion provided by the model. Later the model has been extensively used as a tool for
owners to make portfolio decisions regarding where to invest and divest in industries.
For these reasons, the Five Forces (Porter, 1979) is viewed as a model that explains
both incremental and radical evolution within an industry, but it is not viewed by default
as a model explaining the change for the fundamental rules of value creation. The axis
of value chain vs. external threats poses the first question of whether there is a logical
difference between the two regarding their impacts on industry evolution. In established
industries, the players in the direct value chain might have a long history and
32
knowledge of each other, which would suggest more towards incremental evolution.
The likelihood and impact of external threats is also dependent on the entry barriers of
the industry. According to Porter (1979), the threat of industry includes six major entry
barriers: economies of scale, product differentiation, capital requirement, cost disad-
vantages independent of size, access to distribution channels, and government policy.
Physical production played an important role when the Five Forces was introduced.
Digitalization introduces characteristics that do not fit well into the Five Forces. If the
physical product can be decoupled from the value creation, the model might become
weaker. Digital products and services behave differently than physical products also
with the entry barriers having global scale, real-time product differentiation, low capital,
and virtual distribution. Porter (1979) considered the experience curve to be a major
factor related to the cost disadvantages, which is an important paradigm in several in-
dustries. Porter concluded that the new entrant might be more cost effective than the
experienced competitor when it can utilize a new technology, which is an important
notion in digitalization. Government policies in virtual products and services are related
e.g. to data security and ownership. Afuah and Utterback (1997) combined the view of
the Five Forces and the phases of evolution (fluid, transitional, specific, and disconti-
nuity) and argued that the Five Forces without the context of an evolutionary phase
provides a static view of industry development.
There are also perspectives that consider evolution from a temporal perspective. The
diffusion of innovation describes the process of the adaptation of innovations by cus-
tomers. Rogers (2003) called the phases Innovators, Early Adopters, Early Majority,
Late Majority, and Laggards. In his theory (Rogers, 2003), the size of the population
forms a curve that is similar to normal distribution. Utterback and Abernathy (1975)
concluded that there is a sequence of innovation in which the focus moves from prod-
uct innovation to process innovation after the rate of innovation of the product has de-
clined. The S-Curve is another focal theory that explains the temporality of industry
evolution. Based on the product perspective Foster (1986) suggests that the lifecycle of
a product follows the S-curve where the development is slow in the beginning and in
the end but fast during the time when the product has passed the introduction. The
theory explains that development in the next S-curve is driven by new technology
and/or innovation. The S-curve theory does not implicitly affirm whether the change
affects the dynamics only inside of the existing industry or whether it also applies to
industrial disruption. Klepper (1997) argued that there are a number of industries in
which evolution does not follow the phases of the product lifecycle theories.
33
There are also other perspectives of industry evolution. From the supply perspective,
evolution is often accompanied by scientific discoveries and innovation, whereas mar-
ket-based evolution suggests an unsatisfied or hidden demand. Malerba (2007) divided
the challenges in industry evolution into demand, knowledge, networks, and co-
evolution. He also viewed the evolutionary process as industry-specific. Malerba et al.
(2007) coupled experimental users and demand-driven industry dynamics, and con-
cluded that the evolution of the suppliers of new technology is initially dependent on
experimental users and an unsatisfied niche market.
One of Robertson’s (1967) four conclusions was that innovation can be categorized
depending on the strength of its impact on established patterns. McGahan (2000) sep-
arated industry evolution into non-architectural and architectural, and she divided non-
architectural into Receptive and Blockbuster categories and architectural into Radical
Organic and Intermediating categories. She concluded that the Receptive category is
more stable with a focus on continuous improvement and e.g. distribution, transporta-
tion, and retail would be in that category (McGahan, 2000). The examples are most
likely not static, and McGahan (2000) pointed out that retail was an example of Recep-
tive evolution only before the Web. The Blockbuster category involves large upfront
risks prior to success. A characteristic of both architectural categories is the first mover
advantage, which Radical Organic receives from product and process innovation and
Intermediating receives from innovation related to customers and suppliers.
Low and Abrahamson (1997) added an entrepreneurial view to industry evolution. The
view involves the development phases of the industry in an organizational context. The
phases are Movements (emerging industries), Bandwagons (growth industries), and
Clones (matured industries). The Movement phase is characterized by a need for legit-
imacy, the entrepreneur’s ties to several non-overlapping networks, social factors in
stakeholder motivation, and innovative strategy. The Bandwagon phase rapidly attracts
new competitors, which shifts the focus to resource acquisition. The Clone phase is
characterized by intense competition, and evolution is driven by efficiency of execution.
Disruption in industry evolution implies discontinuity which in business often is attached
to how value is created. Many of the earlier evolution theories contain aspects that can
lead to disruption, but they are not easy to foresee ex-ante; however, disruption seems
to be a phenomenon that is an integral part of an economic system. Schumpeter (2008)
argued that capitalism can never be stationary and that Creative Destruction is the fun-
damental engine that destroys existing economic structures by periodically replacing
them with new ones.
34
Dosi (1982) introduced a theory that argues that demand alone cannot explain industry
discontinuities and that there are also technological paradigms that function similarly to
scientific paradigms. He argued that trajectories are the result of market or technology-
based innovations, whereas discontinuities are a result of a new paradigm that also
considers society at large in addition to the market and technology. This view is similar
to Schumpeter’s (2008) view. Tunzelmann et al. (2008) reviewed technological para-
digms, and 26 years later, their biometric study showed that Dosi’s (1982) argument
had evolved into a focal theory on technological paradigms. They also claimed that
Dosi’s (1982) theory bridged technology-specific knowledge and firm-specific capabili-
ties.
Prescott et al. (2014) defined Industry Convergence as “the blurring of industry bound-
aries that creates competition among firms, which did not compete with another previ-
ously.” Several convergence articles are related to the ICT industry. Lee et al. (2010)
argued that in advanced countries, the competitive advantage in new economies is
highly multidisciplinary and includes scale, scope, expertise, and convergence. They
constructed a framework in which the evolution of convergence and organizational in-
novation had six levels: Product->Functional->Organizational->Technology->Industry-
>Bio-artificial. Hacklin et al. (2010) called the convergence phases as knowledge,
technological, applicational, and industrial phases.
There have been several attempts to design a model that can foresee potential con-
vergence. Fai and Tunzelmann (2001) stated that growing complexity based on patents
drives the flow of knowledge from one industry to another and consequently enables
convergence. Curran et al. (2010) analyzed chemicals, pharmaceuticals, personal care
products, and food and agriculture and concluded that a patent analysis can provide
indications of emerging convergence.
Important contributors to convergence include knowledge flow, competence develop-
ment, and enabling networks. Srinivasan et al. (2007) concluded that vicarious learning
is an important contributor for converging markets by analyzing the data of digital pho-
tography development. Lei (2000) argued that industry convergence places significant
pressure on companies’ core competencies as a competitive advantage and that com-
panies should be more alert regarding the development of adjacent industries. The
importance of networks was studied by Lee (2007), and he used data from the telecom
industry. The framework is shown in Figure 2. His conclusion was that the quality,
quantity, and diversity of the information from networks do contribute to timing and rap-
id entry into the market.
35
Figure 2. Network resources, mechanism, and expected relationships (Lee, 2007)
The European Union called the convergence of telecommunication, information media,
and electronics as TIME industries5. Hacklin et al. (2013) argued that drivers for the
TIME industry are related to the advancement of technology, open architectures and
standards, policy and regulatory reforms, and changes in end-user demands. The view
is slightly different according to an earlier article by Gambardella and Torrisi (1998),
who studied the electronics industry and concluded that technological convergence
does not necessarily imply market convergence, partly due to the large amount of in-
dustry or product-specific assets of the existing actors. The difference in these views
might be explained by the increasing role of intangible assets that can currently be
shared. Benner and Ranganathan (2013) also investigated the telecom industry and
studied investors’ beliefs and analysts’ reactions regarding convergence using data
from VOIP technology. Their conclusion was that divergence works more positively
with growth-driven investors than margin-driven investors. The finding might indicate
inertia in convergence if the investor community behind the potential converging indus-
try is margin-driven. Rao et al. (2006) examined the same industry from a different per-
spective when they studied the fusion of several disruptive technologies and used
Skype as a specific case. They concluded that new business models play an integral
role in technology fusion based on disruptive convergence.
5 http://www. europarl.europa.eu/Jobs of the future require multi-skilling, retrieved 6.6.2016
Quality of Information
Quantity of Information
Diversity of Information
The Timing of Market Entry Brokerage (+)
(early access to more information sources)
Network efficiency ( ) (size vs. cost)
36
2.1.3 Disruptive vs. sustaining change
The heading reflects the earlier viewpoints of evolutionary vs. radical (Abernathy and
Utterback, 1978) innovation or development targeting a current S-curve or a shift to the
next S-curve (Foster, 1987). Christensen’s (2011) book, Innovator’s Dilemma, has
gained vast popularity among practitioners, and it is often referred to when incumbents
discuss the potential impact of digitalization. Schumpeter (2008) and Dosi (1982) dis-
cussed industrial disruption, paradigm changes, and discontinuities, whereas disruption
alone more often implies radical changes in product-based market positions in busi-
ness. This does not exclude a pattern that leads to the disruption of industries. Due to
its merits in day-to-day management practices, this sub-chapter specifically examines
the degree of radicalness using Christensen’s (2011) framework.
The daily activities involved in operating a company and investing in development to
maintain a historical competitive advantage requires a significant amount of time and
effort from management. The fact that the investment pays off might divert manage-
ment’s focus away from activities that target disruptive and riskier opportunities. Chris-
tensen's (2011) Innovator’s Dilemma has gained vast popularity due to its fundamental
argument: doing the right thing might still push you out of business. The fact that the
theory is gaining more popularity now than 20 years ago could imply that there are a
growing number of issues for which the “right thing” has changed; based on this disser-
tation, the change would be enabled by digitalization. The two core elements of his
theory are characterized as follows:
“Sustaining technologies have in common that they improve the performance
of established products along the dimensions of performance that mainstream
customers in major markets have historically valued.”
“Disruptive technologies bring to market a very different value proposition than
had been available previously and underperform established products in main-
stream markets but they have other features a few fringe (and generally new)
customers value.”
Christensen’s framework was based on case research from several industries, i.e.,
computer disks, steel, and excavators. The cases had a commonality that the good
companies provided supplied based on a need expressed by customers, which led to
performance oversupply (more features that customers needed). In these cases, the
disruption was based on a technology that was simpler, less expensive, and reliable.
37
The core elements of his theory are shown in Figure 3, which also illustrates that over
time, disruptive technology addresses the needs of the high-end market.
Figure 3. Impact of sustaining and disruptive technological change (Christensen, 2011)
The low-end market customers are not necessarily the same what Rogers (2003) re-
ferred to Innovators, or experimental users, as defined by Malerba et al. (2007).
Henderson and Clark (1990) developed a framework that examined the potential failure
of incumbents using a concept called Architectural Innovation, which refers to changing
the linkages between the product concept and its core components. If the link remains
unchanged while the concept is reinforced it is called Incremental Innovation, whereas
a discontinuity in the concept refers to Modular Innovation. Changing the link and rein-
forcing the concept is referred to Architectural Innovation and if there is a discontinuity
it would imply Radical Innovation. Henderson and Clark (1990) suggest that “Architec-
tural Innovations destroy the usefulness of architectural knowledge of established
firms.”, whereas Radical Innovation forms a new Dominant Design. Utterback and Ab-
ernathy (1975), who developed the concept of Dominant Design, asserted that there is
a time period in a given industry when a firm can apply a dominant competitive strategy
based on innovative product design.
Danneels (2004) discussed the definition of Christensen’s Disruptive Technology, its
assumptions, and its core conclusions. The overall conclusion was that the theory is
useful but overly simplistic, and consequently, most of conclusions involved the need
Time
Pro
du
ct P
erfo
rman
ce
Performance demanded at the low end of the market or in a new emerging segment
Performance demanded at the high end of the market
38
for future research avenues. Danneels (2004) pointed out unanswered aspects of the
theory, such as:
When does technology become disruptive?
Can ex-ante predictions for firm’s success be made based on theory?
Are there other factors in addition to the incumbents that would explain the dis-
ruptive impact better?
Is there a way to model the customer orientation’s role in disruptive technology?
Is separate activity in disruptive technologies the only successful way to ad-
dress the issue?
Danneels (2004) also argued that the customer-related critique of the theory is not jus-
tified as Christensen’s view was not against customer orientation. Tellis (2006) followed
a similar process in examining the core elements of Christensen’s theory and reached
a conclusion that is similar to Danneels’ conclusion, which is that the theory has be-
come too generic considering the quantity of data it is based on. Tellis (2006) suggest-
ed two specific conclusions:
1. The same problem as with Disruptive Technology occurs in the S-Curve techno-logical evolution.
2. The underlying explanation of visionary leadership is not considered.
Markides’ (2006) critique of Disruptive Technology was that it was too technology-
driven and did not take into account business model innovations, which have different
characteristics. Sood and Tellis (2011) argued that disruptive attacks can occur in both
low-end and high-end markets and that incumbents introduce disruptions as often as
entrants, which are not always less expensive, as Christensen (2011) suggested.
Based on the model that Sood and Tellis (2011) applied to four markets (external light-
ing, desktop printers, data transfer, and analgesics), they reached three conclusions:
1. Disruption takes place from multiple directions.
2. Disruptors do not always introduce disruptive technology.
3. Disruptive technologies do not always follow the S-Curve; the development can
be sporadic.
A recent critique by Bergek et al. (2013) challenged both current sources of incumbent
targeted disruptions, competence (technology) and demand (market), based on case
studies on electric cars and gas turbines. They stated that current industries possess
more inertia than current theories suggest. They explicitly argued that Creative Accu-
39
mulation, a term they used for the concept, requires the simultaneous and rapid fine-
tuning of existing technology, the acquisition of new technologies, and the integration of
those technologies into superior products and solutions. Indirectly, this would mean
that incumbents have better chances to be disruptors than Christensen (2011) sug-
gests, and that sometimes being disrupted might be better explained by the poor im-
plementation by the incumbents than by the Disruptive Technologies theory.
The question of the predictability of Disruptive Technologies was included in Christen-
sen’s critique, and there are number of articles that address this question. Linton (2002)
modeled the factors that have an impact on whether an innovation is disruptive and
used MEMS (Microelectromechanical systems) as a reference, as it has numerous
applications and high anticipated growth rates (lowest 60 % over a 6-year period) with
a high variance (highest 19900 %). He discussed learning curves, changes in market
size, repurchasing, supply constraints, replacement technology, and complementary
products. Based on his quantitative model, he argued that forecasting potential disrup-
tive innovations requires each market to be analyzed independently before reaching a
conclusion. Linton (2002) also asserted that the inhibitors of supply and demand must
be understood to improve the quality of the forecast. Husig et al. (2005) tested Chris-
tensen’s (2011) concept in making ex-ante predictions on the disruptiveness of W-LAN
technology. They concluded that W-LAN would not be disruptive and that the method-
ology is useful for practitioners because it offers a continuous and systematic way of
following the progress of potentially disruptive technologies. Based on their literature
review, Govindrajan and Kopalle (2006) argued that technology-based disruption mod-
els are not effective in making ex-ante predictions for potential disruptive technologies
but that they provide good frameworks for ex-ante predictions regarding which type of
firms could develop disruptive technologies.
Technology strategies and technology roadmaps have been used by management for
future technologies as part of the overall strategy of firms; however, the principle ques-
tion regarding whether or not disruptive technologies can be predicted remains. Walsh
(2004) combined the road mapping tool and disruptiveness, and MEMS was used as a
reference technology. Figure 4 presents the components of the factors needed for
roadmaps.
40
Figure 4. Multitier visual output of a technology roadmap (Walsh, 2004)
The road map illustrates a systematic process based on a framework comprised of
existing theories, expert judgment, and data. The process provided 15 conclusions,
and the framework and process were argued to be useful (Walsh, 2004). The frame-
work includes links between components and systems that were slightly different than
those presented for Architectural Innovation by Henderson and Clark (1990). Walsh
(2004) highlighted the need for a wide coverage of stakeholders in the process to avoid
entrapment by Sustaining Technologies. Kostoff et al. (2004) presented similar findings,
but their view was more process-oriented and also included stakeholder participation
and workshop procedures. They emphasized the potential offered by text mining as a
new technology-based tool (Kostoff et al. 2004).
The lack of competence of the incumbents in relation to Disruptive Technologies is one
critique of Christensen’s model. Henderson and Clark (1990) suggested that Architec-
tural Innovation destroys the existing knowledge of incumbents. Danneels (2004) dis-
cussed the issue indirectly when investigating the need for separate organizations for
Disruptive Technologies. Henderson (2006) argued that incumbents’ weaknesses re-
garding disruption are not only explained by cognitive failures of top management but
also by a lack of technological and market-related competencies. Garrison (2009) also
analyzed the competence perspective of Disruptive Technologies. He concluded that a
response capability is critical in the early adaptation of Disruptive Technologies and
that smaller firms tend to do this better. Garrison (2009) related this response capability
with the technologically opportunistic business culture of a firm.
Christensen has extended the original theory several times since 1997, and in his pa-
per (Christensen et al.) published in 2002, he pointed out that core requirements in
Markets
Systems
Components
Technology
Time
41
applying the theory are the ability to make the distinction between Sustained and Dis-
ruptive technologies or ideas as well as to manage them differently. One of the Chris-
tensen original claim (2011) for incumbents’ weakness to detect disruptive business
ideas is that they appear to be marginal for the faith of the company in the early stage.
Christensen et al. (2002) also discussed the fundamental aspect of repeatability, which
would require a separate process for disruptive ideas, as they are difficult to accom-
plish. They explained that one of the few companies that has provided an array of dis-
ruptive innovations is Sony. Wessel and Christensen’s (2012) recent article included
business models in the disruption discussion, which addressed prior criticism. They
also concluded that inertia limits disruption and outlined five barriers:
Conservative customers
Difficult technological implementation
Lack of proper ecosystems
Immaturity of the technology
An unchangeable current industry cost structure
Wessel and Christensen (2012) posited that the advantage of disruptive innovations
must be realistic, rather than trying to break inertia that is based on natural reasons,
such as ships carrying heavier cargo than planes.
Economy, strategy formation and top management teams 2.2
Stakeholders, and other factors that form the external environment of a firm, have been
discussed in relation to industry evolution. In the value chain, enablers, actors, and
users’ strategic actions also direct the course of future development. The actions are
based on management decisions. This sub-chapter reviews the literature on strategy
formation and top management teams. First, a few perspectives regarding the aspects
of economics in digitalization and knowledge-based businesses are briefly presented.
2.2.1 Knowledge based economy
The fundamental factors of production—labor, raw materials, and capital—will inevita-
bly be involved in machine-building in some form. Firms in the machine-building indus-
try’s value chain also continue to maximize profits to compensate owners through divi-
42
dends and to invest the remaining profits in the firm for future profits. The profit is made
using the factors of production with innovative methods and procedures to deliver
products and services to customers, and the price is dependent on the competition in
the market. Digitalization allows for increasing value, which is delivered in intangible
forms. This is a shift toward a new type of knowledge-based business. Digitalization
allows also knowledge workers to participate in value creation independent of their lo-
cation and at least partly independent of whether they work as private entrepreneurs, in
startup companies, or in multinational corporations.
The World Economic Forum (2015) introduced a concept of Output Economy, that uti-
lizes smart connected devices, and as a result, enables optimization on a higher sys-
temic level, whether it involves energy saving or improved health. This also expands
the concept of products and services. The Outcome Economy concept is illustrated in
Figure 5. The concept was presented as a roadmap to the Industrial Internet, but it also
reveals that the higher the development stage, the less dependent the value creation is
on the original factors of production, except of the knowledge workers if they would be
defined as labor in the original context.
Figure 5. Outcome Economy concept (World Economic Forum, 2015)
1.Operational Efficiency • Asset utilization
• Operational cost reduction
• Worker productivity
2. New Products & Services
• Pay-per-use
• Software-based services • Data monetization
3. Outcome Economy • Pay-per-outcome
• New connected eco-systems
• Platform-enabled marketplace
4. Autonomous Pull Economy • Continuous demand-
sensing
• End-to-end automation
• Resource optimization & waste reduction
43
Because an increasing amount of value in digitalization is based on processing
knowledge, a resource-based view on economics is crucial. Conner (1991) studied the
topic in the context of an industrial organization economy and identified five schools of
economy: Perfect Competition, Bain, Schumpeterian, Chicago Responses, and Trans-
action Cost. By analyzing those against what was known on resource dependency, she
concluded that:
“Performance differentials between firms depend to significant measure on pos-
session of unique inputs and capabilities.”
She highlighted the importance of capabilities in collaboration with Prahalad (Conner
and Prahalad, 1996) by arguing that privately held knowledge is a focal source of the
competitive advantage of a firm. In the digital era, both privately held resources and
transaction-based fundamentals are challenged. Crowd concepts push the limits of
firm-centric resources.
The transaction economy is also challenged by a sharing economy, or open source
movement, in which contributors might have hedonistic, intellectual, social, or other
similar motives that cannot be explained by rational economic theories. In his book
Managing in the Next Society, Peter Drucker (2002) emphasized cognitive science and
posed a question to firms: how are knowledge workers attracted? A mismatch of firms’
values has resulted in some knowledge workers remaining independent and applying
their own values in open networks, which was the case in the Linux movement.
Uber and Airbnb popularized the sharing economy discussion based on the financial
benefits of asset sharing. Belk (2010) studied sharing and made a distinction between
sharing, gift giving, and commodity exchange. He also argued that a powerful driver in
a sharing economy is the changing attitude toward ownership. Sharing in B2B is not
new; shipping lines have shared vessel capacity between competitors for years; how-
ever, sharing in B2B is more similar to a transaction economy, as it is based on con-
tracts that reserve capacity. For consumers, sharing seems to be driven by platform
businesses that act as marketplaces that contain the information needed for sharing to
be effective. Ownership issues were also examined when Bardhi and Eckhardt (2012)
investigated sharing in the context of passenger cars in the Zipcar case. They defined
Access Based Consumption as “transactions that can be market mediated but where
no transfer of ownership takes place” (Bardhi and Eckhardt, 2012). They also conclud-
ed that market-mediated sharing eliminates the social reciprocity aspect. John (2013)
analyzed the rise of sharing Web 2.0 not only from an economics viewpoint but also
44
based on its societal impact. Sharing ideas, which is enabled by information networks,
is also a growing phenomenon. The platforms are reviewed in sub-chapter 2.3.1, and
the open innovation paradigm is reviewed in sub-chapter 2.3.2.
A final note on economics is that the division for B2B and B2C is also blurred by indus-
try convergence. Fifteen years ago, Achrol and Kotler predicted (1999) that a network
economy would have similar consequences for B2B as it has had for B2C.
2.2.2 Strategy formation
Firms should answer the type of research questions of this dissertation as part of their
strategy formation. Whether digitalization will enable industrial disruption should be
determined based on that process, and whether to ignore, adapt, or become a disrup-
tor should be determined based on external opportunities and internal capabilities.
Management typically uses explicit tools in its strategy formation, or at least post-ante it
can likely be concluded that there was a certain type of strategy thinking utilized to
reach conclusions and make decisions.
The tools are often derived from particular strategic thinking or strategic schools, and
therefore they logically reduce the perspective to that of the premises of the strategy
tool.
Business environment is not stable, which would imply that schools and tools are also
to a certain degree of temporal. The selection of tools might also impact the actual
strategic choices. In this dissertation, the schools and tools are reviewed based on po-
tential disruption enabled by digital technologies. Value chain and process thinking,
industry beliefs, and business models are reviewed in this context in their own sub-
chapters.
Strategy can be analyzed from a corporate or a business perspective. Corporations
often approach the strategy from an owner’s perspective, including visions, missions,
values, portfolios (which businesses they are in), financial structure, performance man-
agement and organizational structures. If the corporation operates in multiple busi-
nesses, they have to design the roles between the corporate and businesses as well as
the themes they wish to drive across businesses. Corporations must prove that they
are more than the sum of their parts. Collis and Montgomery (1998) suggested that
resource allocation (general vs. specialized) can be utilized to analyze this, which takes
45
into consideration the scope of the business, coordination mechanisms, control sys-
tems, and the size of the corporate office. Businesses must make choices regarding
how they create value by serving their customers better than the competition, which
often includes decisions on markets, customers, products, and services. In addition,
cost vs. value, required capabilities, and numerous functional statements, such as dis-
tribution and production models (owned or outsourced), should be considered. Strategy
is also a process that involves a varying number of participants of the organization or
the participation of stakeholders. Minzberg and Lampel (1999) discussed the evolu-
tionary nature of the process due to learning, borrowing, and blending new with old in
the competition in evaluating new ideas. It is also known that planning without commu-
nication and implementation does not yield the expected results.
There are number of articles and books that have discussed the different perspectives
or schools of strategy. Several of the schools evolved based on a theory of a scholar or
a framework of practitioners. Minzberg et al. (1998) listed 10 strategic schools: Design,
Planning, Positioning, Entrepreneurial, Cognitive, Learning, Power, Cultural, Environ-
ment, and Configuration. Cummins and Wilson (2003) distinguished 13 different per-
spective of strategy: Ethos, Organizing, Intention & Anticipation, Orchestrating
Knowledge, Data and Sense-Making, Creativity, Exploration & Interconnection, Sys-
tems Thinking, Process-Power-Change, Marketing, Numbers, Decision Making, and
Orientation & Animation. De Wit and Meyer (2005) framed strategy into Process, Con-
tent, and Context. Content includes strategic thinking, formation, and change. Näsi and
Aunola (2002) used a framework in which strategic logic included four perspectives:
Strategic Management and Strategic Management of Structures are administrative in
nature, whereas Strategic Game and Strategic Leadership are situational and environ-
ment-driven.
The 10 schools by Minzberg et al. (1998) comprise a comprehensive framework that
combines strategic thinking and strategy formation, and it is used here as a framework
to review respective theories. A number of strategic tools are placed inside of the
schools and some of them are developed later than Minzberg et al. (1998) published
their framework.
The Design School requires understanding the external environment in relation to a
firm’s internal capabilities, and therefore the SWOT (internal strengths and weaknesses
and external opportunities and threats) tool is included in this school. This school also
highlights the architectural and conceptual character rather than the form, as the name
suggests. Humphrey (2005), one of the developers of SWOT, underlined the im-
46
portance of formal planning when using the tool in his retrospective description. Minz-
berg et al. (1998) argue that the Design School has been one of the most used one by
practitioners in strategy formation. Chandler’s assertion (1990) that structure follows
strategy fits to the philosophy of this school. Case methodology has a strong link with
this school as well.
The Planning School emphasizes the formal planning process. Planning as a discipline
is familiar in engineering or product development in which structures, hierarchies, pro-
cesses, checklists, iterations, and reviews are used daily. If top management has a
focal role in the Design School, detailed planning implies a greater role for planners,
and the management role shifts toward an approval role. Minzberg et al. (1998) refer
Igor Ansoff as one of the important pioneers in this type of strategic thinking.
The Positioning School assumes that based on analyses, an optimal strategy with ge-
neric characteristics can be chosen. BCG matrix is a prime example of that. Using the
BCG matrix (Henderson, 1970), businesses are analyzed based on the growth of the
market in relation to the market position, and depending on the status of the business
in the portfolio (matrix), there is a specific optimal generic strategy available. Later Por-
ter’s (1985) generic strategies of cost focus or cost leadership and differentiation or
differentiation focus became almost the industry standard. Ansoff’s product / market
window (Ansoff, 1957) is however a good example of a tool that can be applied with
logic of Design, Planning or Positioning School. In that model product or market could
be either existing or new and the model was designed for diversification strategies. The
analysis, market attractiveness, and choice of generic options phases comprise an
integral part of this model, and therefore it can be argued that it belongs to this school.
The bestseller Must-Win Battles (Killing et al., 2006) can be included here due to its
analytical process coupled with generic advice, although it has characteristics of the
Design School (external opportunities vs. internal capabilities).
The Entrepreneurial School is based on visionary leadership, and according to Minz-
berg et al. (1998), it can also be defined as strategic thinking by seeing through (behind,
above, below, beyond, ahead, and beside). They also connected this to the Schumpet-
erian (2008) view of entrepreneurialship in economics. This type of person-centric and
intuitive strategy formation has been linked to corporate leaders, such as Jack Welsh,
and owners of entrepreneurial companies, such as Richard Branson of Virgin and Ste-
ve Jobs of Apple. In the digital domain, Eric Ries’ Lean Startup (2011) has become an
important handbook. The attributes of entrepreneurial management and intuition in the
development of this handbook could place concepts such as the Minimum Viable
47
Product in the Entrepreneurial School, but Ries’ (2011) focal element of continuous
learning would suggest that the Learning School would be an equally justified interpre-
tation.
The Cognitive School is different from the other schools because it focuses on strate-
gists’ mental processes. The role of the Cognitive School in this dissertation is related
to management cognition as one sub-driver that can explain the inertia or acceleration
of drivers that cause industrial disruption. Therefore, management cognition is re-
viewed in more detail in sub-chapter 2.2.3.
The Learning School involves the connotation of capabilities and strategy formation.
The Core Competence of the Corporation article by Prahalad and Hamel (1990) is one
milestone of that school. They defined Core Competencies as “collective learning in the
organization” in integrating diverse production skills or multiple technologies as well as
organizational value creation. This implies that resource development is a vital task of
top management. Conner and Prahalad (1996) shared a similar view regarding re-
source dependency in a knowledge-based economy. Minzberg et al. (1998) included
the perspective of rule breaking and ignoring industry beliefs as characteristics of the
Learning School, although this visionary attribute could also be included in the Entre-
preneurial School. Rule breaking is mentioned in the principles of Hamel’s (1996) strat-
egy formation: “If there is any hope of industrial revolution, the creators of strategy
must cast off industrial conventions.” The same concept was discussed in Kim and
Mauborgne’s (2005) Blue Ocean Strategy (reconstructing market boundaries) and the
Business Model Generation by Osterwalder and Pigneur (2010). Unlike Hamel (1996),
who emphasized resources, Kim and Mauborgne (2005) and Osterwalder and Pigneur
(2010) both highlighted the process. Osterwalder and Pigneur (2010) developed later
their thinking into the Business Model Canvas, which is their trademark in this decade’s
strategy tools.
The Power School suggests that strategy formation is a political process. Although
there are elements of political compromises in the strategic processes of firms, perhaps
with the exception of the Entrepreneurial School, this school emphasizes observation
rather than a set of tools for practitioners. Due to the external compromises, the school
is related to the stakeholder theory (Freeman, 2015). Minzberg et al. (1998) discussed
the role of strategic alliances in this school; however, it can be argued that forming alli-
ances is one strategic choice regardless of whether or not the risk of compromises in-
creases.
48
Minzberg et al. (2010) stated that the Cultural School is based on the concept that
strategy formation is also a social process. For effective results, the cultural aspects of
participants must be understood and considered. For competence or resource-based
views, there is a stronger need to manage the cultural dimension as well, as discussed
previously; however, it could be argued that managing cultural dimension is not only
relevant in strategy formation but also in all leadership or stakeholder relations. The
national culture poses a dimension of its own in the international business environment.
The cultural aspect is related to this investigation in the same way that management
cognition is related, which will be reviewed in sub-chapter 2.2.3.
The Environmental School reduces strategy formation to a responsive role. The fun-
damental belief is that there are a number of general forces that affect strategy for-
mation; however, firms cannot influence them. Minzberg et al. (2010) listed population
ecology as one of the main sources of strategic thinking. During the investigation of this
dissertation, it was assumed that digitalization is a systemic phenomenon that has an
impact on knowledge creation. If this assumption is correct, then no firm or group of
firms could change it. As a critique to a passive response, the premise would still be
generic enough to allow for successful strategy formation.
The Configuration School is a temporal dimension of the schools. It presents different
models of organization based on the premise that their use is situation-driven, which is
similar to Chandler’s (1990) assertion. As moving from one organizational ideal to an-
other is a laborious process, Minzberg et al. (1990) presented several examples re-
garding transformation or change management for this school. Thus, in this dissertation,
the aspects of the Configuration School are similar to management and leadership,
which are reviewed in sub-chapter 2.2.3.
Minzberg et al. (1990) classified their schools in two different ways. First, they classi-
fied Planning, Design, and Positioning Schools as Prescriptive Schools, and the re-
mainders were classified as Descriptive Schools. Then, they illustrated the activities
based on publications and management attention over time. The Prescriptive Schools
had a medium level of support due to the introduction phase in the early 1970s. The
Planning School was popular from the mid-1970s to the mid-1980s when the Position
School gained popularity over the declining Planning School. The Descriptive Schools
have had growth activity since their respective introductions, but the Learning and Con-
figuration Schools experienced increased growth. The Configuration School might not
be comparable because it is considered to be a temporal hybrid of the other schools.
Another synthesis method involves placing the schools in a matrix where the predicta-
49
bility of the external environment is the y-axis, and the x-axis represents the rationality
of the internal process. The synthesis is presented in Appendix 1, and the main conclu-
sion is that the extreme unpredictability and natural internal processes are characteris-
tics of the Cognitive and Learning Schools.
The analysis of strategic schools by Minzberg et al. (2010) did not discuss all major
tools used by practitioners. All schools have some type of implicit or explicit assess-
ment as part of the strategy formation, and tools such as the Key Success Factor
based analysis and planning by Daniel (1961) or the popular Balanced Scorecard
(Kaplan and Norton 1992), which combined customer perspectives, business objec-
tives, and development plans with financial measurement, played or still play a role.
Also, specific frameworks for holistic strategies, such as 7S (Structure, Systems, Style,
Staff, Skills, Strategy, and Superordinate Goals) by Waterman et al. (1980) have been
useful without being strongly linked to a certain school, which is the case also with pur-
pose-built models, such as the Repeatable Model (Zook, Allen 2012), for exploring
growth opportunities based on the current core of a firm. Rigby and Bilodeau (2005)
conducted a management tool survey and found that tools are temporal and have dif-
ferent strengths and weaknesses, and they also recommended that tools be adapted to
the situation of the firm. This would imply that the tool selection can affect the outcome
negatively if the incorrect tool is chosen. The same study (Rigby and Bilodeau 2005)
revealed that at the time of the research, innovation was a vital element of corporate
success but tools for senior management were not up to par, whereas the most effec-
tive tools were available for strategic planning.
This dissertation investigates disruption or paradigm changes in lieu of quantum leaps
in market positions or continuous improvement. Digitalization is described as a system-
ic phenomenon. This would locate the enquiry in the place of the framework (Appendix
1) where the external world is uncertain and complex, and the internal process is closer
to natural than rational. Based on the framework, this would suggest closeness to the
Learning School, and that cognitive and cultural factors should be considered.
2.2.3 Top management team
Management plays a central role in strategy formation, which was discussed in the
previous sub-chapter, and it plays an equally important role in the implementation of
strategies. This sub-chapter focuses on the soft perspectives that might impact strategy
50
formation and implementations, such as culture, values, and leadership, as well as
more tangible aspects, such as structure and governance. If the assumption of com-
plexity and uncertainty is correct, strategic choices do carry more weight. Was Kodak’s
failure to see digital photography (Lucas and Koh, 2009) linked to culture and leader-
ship, or is Christensen’s (2011) Innovator’s Dilemma in which success in sustaining
technologies prevents incumbents from identifying emerging disruptive technologies
connected to the mental models of management and organizations?
Culture and values
There are cultural aspects linked to firms as well as to the society or country to which
the members of those firms belong. The same applies to the values of the firms’ vis-à-
vis the values of the individuals who work for those firms. Schein (2009) defined culture
as “a pattern of shared tacit assumptions that was learned by a group as it solved its
problems of external adaptation and internal integration that has worked well enough to
be considered valid and, therefore, to be taught to new members as the way to per-
ceive, think, and feel in relations to those problems.” The practical test for the inquiry of
disruption or paradigm change is whether a firm is risk adverse or not. Hofstede et al.
(2010) defined culture as a formation of stacked layers in which the inner layer is val-
ues and following layers are rituals, heroes and symbols. Practices cut the stacks
across and they are visible to external observers, but their cultural meaning is only un-
derstood by the insiders.
It is rare that multinational corporations have monolithic cultures, as they are often
formed by acquisitions, and thus they inherit the cultures of the acquired companies in
addition to the variances of local cultures and customs of the countries and locations in
which business entities operate. The influence of top management on culture is ques-
tionable. A strong wave of corporate culture took place at the same time that Japanese
companies began to succeed globally based on quality, which was not only a result of
methods but also of a strong corporate culture of quality. For this dissertation, the cul-
tural question is partly related to corporate comfort zones. Schein (1984) pointed out
that anxiety reduction in corporations is a more stable phenomenon than positive prob-
lem solving with innovation, which is a fundamental part of human nature. Hofstede et
al. (2010) identified four relevant business dimensions that had cultural differences:
power distance, collectivism versus individualism, femininity versus masculinity, and
uncertainty avoidance. The premise of the movement was that culture was stronger
than written strategy and that there was a causal relation between corporate culture
and performance. Barney (1986) concluded that organizational culture can be a sus-
51
tainable competitive advantage if it is valuable, rare, and difficult to imitate. Corporate
culture was harnessed in the late 1980s and early 1990s as a management tool to
build a desirable culture. This took place at a time when globalization accelerated, and
corporations began to increasingly deal with issues of managing people with diverse
cultural backgrounds. Schein (2009) pointed out the importance of subcultures, which
stem from mergers and acquisitions, globalization, technological complexity (occupa-
tional cultures), and opportunities offered by information technology.
Based on their research, Tellis et al. (2009) stated that the flattening world has dimin-
ished national factors on radical innovation, but the importance of internal corporate
culture is strong. Recently, Google’s attempt to maintain a start-up type of innovation
culture by allowing employees to devote 20 % of their work time to individual projects is
a well-known example. The top management vehicle often involves a change program
towards the desired culture, whether it is an innovation, performance, quality, or safety
culture. The time required to accomplish this often exceeds the allotted time of the pro-
gram. In the midst of the corporate culture boom, Beer et al. (1990) discussed the rea-
son that change programs do not produce change. Their conclusion was that the ma-
jority of the programs began from the top and targeted changes in attitudes, whereas
the focus should have been on tasks, roles, responsibilities, and learning.
Hofstede et al. (2010) defined values as “broad tendencies to prefer certain states of
affairs over others and continued by stating that values are also feelings with an added
arrow indicating a plus or a minus side and deal with pairings like moral versus immor-
al.”
Based on the literature review, Schwartz and Bilsky (1987) stated that the definition of
values has five common features:
“Values are (a) concepts or beliefs, (b) about desirable end states or behaviors,
(c) that transcend specific situations, (d) guide selection or evaluation and (e) are
ordered by relative importance.”
As with culture, there is a justified question regarding the extent to which top manage-
ment can impact the values of a firm as well as how quickly it can be accomplished. In
family-owned companies, the owners logically have more impact and say regarding
values. Value definition played a similar role in firms as corporate culture but more than
10 years later. While the initial proposition was to create an innovative corporate cul-
ture, innovation was eventually listed as a corporate value. A similar challenge in multi-
52
national companies also occurred: corporations might be formed as a result of multiple
acquisitions, and they operate in several countries. One can separate two extremes
when defining the company values: synthetic aspirational values vs. defining underlying
values based on the company’s evolution. A detrimental case occurs when the written
values contradict the behavioral values. One example is public punishment for mis-
takes when the firm claims that experimentation is a core value. The role of culture and
values in this dissertation is a potential sub-driver for some of the forces that drive or
inhibit industrial disruption that is enabled by digitalization, and an experimental culture
is an example of an accelerative sub-driver.
Structure and governance
Structure has been briefly discussed in strategy formation. In large firms, corporate
strategy often defines the structure and governance as well as the role of the corporate
office, as stated by Collis and Montgomery (1998). The strategy schools include differ-
ing views of the role of structure. “Structure follows strategy” by Chandler (1990) is a
different concept from the resource-based strategy scholars use. In all cases, the board
of directors and top management also play a formal role in the firm, which is governed
by company laws, and the roles vary depending on whether the firm is publicly or pri-
vately owned.
The size of a firm and its complexity most likely has an impact on the optimal govern-
ance model. There are also managerial limits that impact performance. Penrose (1995)
discussed this in his book, and his view can be interpreted as arguing that the limit is
only temporary, as top management ultimately determines its organization’s structure
and who it wishes to hire based on experience, competence, or personality. Booz-
Allen and Hamilton (Kocourek et al. 2000) offered corporations four different types of
governance choices: holding, strategic management, active management, and opera-
tionally involved.
The organization design to create value in firms varies. Line vs. staff is a familiar con-
cept in the military; functions in firms’ are normally discipline-based. Process and ma-
trix organizations are both hybrid versions of a straight forward line organization. Net-
work economy has blurred the boundaries of internal and external organizations. Or-
ganizations are also normally hierarchical and divided into groups, divisions, lines, de-
partments, and teams, and the extent of the hierarchy correlates with the size of the
organization. The type of business (consumer, industrial, service, global, local, project,
53
and multi vs. single product) that the firm is in, also impacts top management’s optimal
choices regarding how to organize.
The examples presented are practical examples to address the question regarding how
firms should be organized rather than an exhaustive review of organizational theories.
The question that is related to this investigation is whether organizational preferences
are connected to potential disruption. Based on research by Charitou and Markides
(2003), 68 % of firms that engaged in disruptive strategic innovation activities did so in
a separate unit. Corporate labs, for example, normally do not have the responsibility of
commercialization. In the current millennium, and specifically due to the unprecedented
success of startups, there is a new renaissance for corporate ventures. The Finnish
ICT Corporation Tieto is a prime example. This corporation established an internal
startup in 2014, and the CEO Kimmo Alkio discussed his rationale:
"The objective is to break boundaries, change operational clock speed, get closer to customers and speed up decision making.”6
Establishing a separate organization with disruptive activities implies that there is an
assumption of a potential disruption or that the firm wishes to be a disruptor. Using an
internal organization or external networks relate to the operational model of the firm.
Use of external networks, which are discussed in the sub-chapters 2.3.1, Platforms,
and 2.3.2, Open Innovation, can also include a notion that external networks can yield
more novel ideas than the internal resources of the firm is able to provide.
Leadership
Visionary leadership is a characteristic of the Entrepreneurial School of Minzberg et al.
(2010). An organization can have a leadership culture, and managers can be referred
to as great leaders, such as Steve Jobs, though his leadership style was criticized.
Leaders manage other people and therefore have an impact on firms or society at large.
Leadership was also discussed by Sun Tzu (1963) or Machiavelli (2003). The Oxford
Dictionary generically defines leadership as “the action of leading a group of people or
an organization, or the ability to do this,7 which connects leadership to personal compe-
tencies. Large organizations typically train management in leadership. Perhaps the
most widely used equilibrium of leadership styles in business is autocratic vs. demo-
6 Tieto CEO Kimmo Alkio in Kauppalehti 23.2.2015
7 http://www.oxforddictionaries.com/definition/english/leadership, retrieved 10.6.2016
54
cratic. Tannenbaum and Schmidt (1958) discussed this equilibrium as well as the need
for situational aspects of the styles. They concluded that the difficulty of right style in
right situation lies in the inner forces of the leader: the value system, confidence in
subordinates, personal leadership inclinations, and feelings of security in an uncertain
situation.
There are several other typologies available, and Westley and Minzberg (1989) divided
visionary leadership into creator, proselytizer, idealist, bricoleur, and diviner. The vi-
sionary attribute is often related to the novelty of the actual strategic choice, whereas
leadership styles can be related to the way that the strategy is implemented. Child
(1972) discussed leadership aspect based on his dominant coalition concept and its
power over the structural design of an organization when making strategic choices. A
dominant coalition generally refers to top management, but in extreme cases, it can
refer solely to the leader. In a turbulent environment, leaders manage uncertainties by
simultaneously implementing incremental improvement and transformation (Tushman,
1997).
Deschamps (2005) analyzed the necessary leadership skills in the context of the type
of innovation strategy applied by a firm. The model is presented in Figure 6, and trans-
formative innovations, that digitalization might enable, can be placed in the top right
corner, as it is systemic and often includes external partners. According to Descamps,
this leadership style refers to a “pragmatic architect” who envisions a new system and
also understands the components and their relations as well as how to accomplish the
goal when some of the critical knowledge is outside of the firm’s borders. Moore (2005)
defined innovation categories based on product leadership, customer intimacy, opera-
tional excellence, and category renewal. He also pointed out that the required leader-
ship style varies based on the category. Moore (2005) expanded the thinking into two
categories: the core – your unique source of competitive advantage and context–
everything else. He stated that core becomes context over time, and the benefit of a
startup is that the focus is predominantly on the core. He also stated that context cre-
ates harmful inertia for renewal, and consequently, the leadership challenge is to rec-
ognize and prevent it.
55
Figure 6. Typology of innovation by strategic focus (Deschamps, 2005)
Managerial Cognition
In strategy formation, cognition is a characteristic of the Cognitive School by Minzberg
et al. (2010). It was concluded that it does not provide an explicit view on strategy but
rather a group of theories that explain management decision making. This is particular-
ly important if it is assumed that during emerging disruption or paradigm change, deci-
sion making is more difficult. One of the salient theories that explain management
choices was presented by Kahneman and Tversky (1979), who challenged the utility
theory’s assumption that gains and losses work the same way depending on the state
of wealth by introducing the Prospects Theory. Empirical data showed that utility curve
is steeper for losses than for gains. Kahneman and Tversky (1979) introduced a
weighting function to the formula that includes both probability and desirability; they
also concluded that with very low probabilities, events are normally ignored, whereas
very high-probability events are regarded as certain. The reference point also plays a
major role in the context of digitalization. For instance, firm A would base their deci-
sions on the actions of main competitors B, C, and D, and a small gain is perceived as
a victory, even though it would fall behind the digital development of company X in an
adjacent industry.
Totally New Product Category
Totally New
Service Offering
Totally New Business Model
Totally New
‘System Business’
New/Improved Product or Process
New/Improved Service Offering
New/Improved Customer ‘Solution’
New/Improved
Customer System
New business Creation
Business Reinforcement
WHY? (OBJECTIVE)
Product/Service ‘System’
WHERE (SCOPE/FOCUS)
Radical
Incremental
Internallya With others
b
HOW MUCH? (INTESITY)
aIncl. with outsourced- or
partners’ technology bTypically with external
‘complementors’
56
Intuition has been studied extensively in decision-making theories, and it can be as-
sumed that strategic choices are common decisions for senior executives. Simon (1987)
stated that in a management context, intuition is similar to professional judgment, and
intuitive decision-making seldom takes place without an analytical process. Kahneman
(2011) discussed the dilemma of intuition vs. analytical / rational in the context of fast
and slow thinking using a vast number of examples, such as hindsight inadvertently
distorting objectiveness. Another important contributor is framing, and Kahneman
(2011) illustrated that a factually identical situation might lead to a different judgment
based on framing. Rosenzweig (2007) argued that causality, which is often simplified,
is an important source of management delusion.
According to Webster,8 the world view is “a comprehensive view of the world and hu-
man life,” and it could be argued that a world view of digitalization is still developing.
Most likely, there are already strong “world views” that might be diverse or even polar-
ized. This could also be the case in the long-term; Starbuck (1982) stated that no or-
ganization has a complete monolithic ideology, and even individuals can have multiple
ideologies and still act logically. Smircich and Stubbart (1985) discussed whether the
interpretation of the environment by a strategist is objective or perceived, a concept of
the Enacted World, and they stated that “environments are enacted through the social
construction and interaction processes of organized actors.”
Simon (1979) viewed rational human behavior as subject to major boundaries of the
environment as well as to the capacity of human beings (Bounded Rationality). Later,
Simon (1991) also pointed out that the same mechanisms are relevant for organiza-
tional learning. One factor that reduces the rationality according to Kahneman (2003) is
Judgment Heuristics, which causes the decision-maker to use a limited number of heu-
ristic principles when solving complex problems. According to Kahneman, this is useful
but can lead to systematic errors.
Stubbart (1989) divided managerial cognition into categories, networks, and inferences.
He argued that the categories included experiences that are prerequisites for human
cognitive abilities. Strategic schools of the previous sub-chapter serve as a good ex-
ample of such categorizing. Stubbart (1989) also argued that the categories used by
managers differ from those used in management science and also differ between man-
agers. One of the issues raised was that managers’ cognitive maps are flawed due to
8 http://www.webster-dictionary.org/definition/worldview, retrieved 15.7.2015
57
excess causality or because they are based on strategic concepts that are too abstract.
In Stubbart’s (1989) article, managerial inference was based on rationality, heuristics,
and expertise; he used integrating of information from a SWOT analysis to strategy
formulation as an example of a cognition problem.
2.2.4 Value chain and process thinking
Value chains play a pivotal role in this dissertation from three perspectives. First, they
are unbuilt into the research question where potential industry disruption and its drivers
are analyzed separately based on the view of enablers, actors and users. Second, the
value chain plays a central role in Porter’s Five Forces (1979), which is a seminal theo-
ry of industry dynamics. Third, industrial companies have included value chain per-
spective as part of their strategies, operations, and development activities for decades.
The value chain is another popular concept developed by Michael Porter that yielded a
large number of citations even 30 years after it was introduced in his book Competitive
Advantage—Creating and Sustaining Superior Performance (1985). Figure 7 describes
his concept in which directly added value activities are called “primary activities,” and
activities that enable the direct activities are called “support activities.” If value chain
thinking is a successful concept in operational improvements in businesses, then can
the same concept create inertia in a potential emerging paradigm change?
Figure 7. Value chain (Porter, 1985)
FIRM INFRASTRUCTURE
HUMAN RESOURCES MANAGEMENT
TECHNOLOGY DEVELOPMENT
PROCUREMENT
INBOUND LOGISTICS
OPERATIONS OUTBOUND LOGISTICS
MARKETING& SALES
SERVICE
PRIMARY ACTIVITIES
SUPPORT ACTIVITIES
58
Porter’s value chain (1985) was publicized at the same time that total quality manage-
ment began to flourish. The English translation of Ishikawa’s (1985) What is Total Qual-
ity Control? The Japanese Way was a milestone in which the philosophy of breaking
functional silos and focusing on processes took place (“the next process is your cus-
tomer”). Value chain thinking was also rooted in the business process of re-engineering
and process management. Hammer stated (1990) that one must rethink the way differ-
ent phases are completed in the process. In essence, it calls for a quantum leap in pro-
cesses but very seldom is a source of renewal in company strategies or at the industri-
al level; however, there are notable examples of supply chain innovations that changed
the industry practices, such as the well-documented Dell case in the computer industry.
Lee (2002) commented that Dell’s success was based on their highly modular product
structure, their build-to-order concept, and the ability to maintain a stable process in a
highly unpredictable demand of innovative products. Process management was revital-
ized due to large enterprise systems, as consultants introduced a practice in IT projects
in which the first phase was a description of a current process (“as is”) followed by an
aspirational “to be” process. This practice included an intrinsic logic of improvement.
In the development work of a firm, value chains and process thinking are combined
with decision making in Cooper’s gate model. Cooper (1990) stated cycle time reduc-
tion and improvement of hit ratios as drivers of the philosophy of gate systems in inno-
vations. For example, the project can only progress to the next phase after the gate
approval, but the goal of the system was a stronger market analysis in the early phases.
Critique for the gate model involves the easy termination of a project, especially in
large organizations in which there are large numbers of staff resources to evaluate
proposals. This can proportionally increase incremental innovation that has a higher
success rate but a smaller upside. Veryzer (1998) contended that the gate model is not
the optimal approach to discontinuous innovation. The software industry challenged the
sequential concept of gate systems in development for the same reasons that Cooper
originally presented—time and hit ratio—by introducing the agile manifesto9 in 2001,
which included the following principles:
Individuals and interactions over processes and tools
Working software over comprehensive documentation
Customer collaboration over contract negotiation
9 http://agilemanifesto.org/ retrieved 10.6.2016
59
Responding to change over following a plan
The manifesto led to various methods, such as Scrum. The unique feature of Scrum
(Schwaber and Beedle, 2002) compared to the gate model is Sprint, which is a prede-
termined period that is less than four weeks. The Scrum team develops a product or
feature in the Sprint that meets the requirements and is ready for publishing. Another
unique feature of Scrum is the retrospective phase, which is a session to evaluate the
work of the team and present improvement ideas before the next sprint. Sprint retro-
spective can last a maximum of three hours.
Recently, several radical value chain and process innovations have been enabled by
digital technologies, and some were coupled with a business model change. This has
occurred both in primary and in support activities or processes. Digital entertainment is
a good example in which the process of satisfying a need, such as listening music, has
been reinvented—downloading files instead of buying records or CDs from a shop. If a
firm can deliver their products or services through digitalization, legacy systems are still
a challenge. Incumbents often invest a vast amount of money in IT systems that make
their current processes effective. Writing them down or upgrading them to accommo-
date a new process or business model is expensive. Often, the new type of infrastruc-
ture is service, which is attractive from a capital expenditure point of view, but the
switching cost is a critical issue. Trigeorgis (1993) discussed this from a financial man-
agement perspective and determined that there should be financial modelling for active
management to change the conditions of the original investment, which are IT systems
in this case. For disruptive platform investments, the problem is twofold if an incumbent
wishes to be the disruptor. First, the incumbent is responsible for the cost of legacy
systems, and at the same time, it should invest in the possibility that the market would
follow, including competitors, which resembles the Prisoners Dilemma10 as a decision
problem (In game theory, a situation in which two players each have two options whose
outcome depends crucially on the simultaneous choice made by the other, often formu-
lated in terms of two prisoners separately deciding whether to confess to a crime).
Cable and Shane (1997) studied the Prisoners Dilemma in relation to venture capital-
ists and entrepreneurs and concluded that cooperation is a preferred option. This could
be interpreted in a value chain context that two important competitors could cooperate
in building a new “standard” for the value chain for the benefit of the industry.
10 http://www.oxforddictionaries.com/definition/english/prisoner's-dilemma
60
There has been a tendency in high-wage countries to focus activities on the users,
leave the upstream to suppliers, and act as a system integrator. This is evident in com-
panies that manufacture the hardware of digital technologies. Therefore, it is notable
that successful digital companies also practice different value chain strategies, such as
Tesla Motors, which is a small car company with a business model that differs from the
models of the automobile industry. At the same time, they have a high integration in
production, own key components, and have a massive manufacturing site11 despite
their digital business concept.
2.2.5 Industry beliefs
Incumbents operate in established industries. Are there recipes that work collectively
similarly to management cognition on an individual level? An industry recipe allows
management to identify a sound space for their strategy that includes technological,
financial, and social dimensions (Grinyer and Spender, 1979). Examples include econ-
omy of scale, need of flexibility, local vs. global, and asset or research intensive. An-
other example is the early mover advantage, which has a link to learning. Spence
(1981) concluded that learning can create similar entry barriers as economies of scale
in certain industries. Based on extensive case studies, Spender (1989) stated that the
dynamics of an industry recipe are not driven by what firms believe about the industry
but what managers believe about their companies.
Dominant Logic is similar to an industry recipe. Prahalad and Bettis (1986) defined it:
“a dominant general management logic is the way in which managers conceptu-alize the business and make critical resource allocation decisions, be it technolo-gies, product development, distribution, advertising or human resource manage-ment.”
They also argued that the limit of diversity of corporations is not driven by the variety of
markets or technology but by the variety of dominant logics. Their suggested sources
of Dominant Logic are shown in Figure 8.
11 http://www.teslamotors.com/gigafactory retrieved 7.7.2015
61
Figure 8. Sources of Dominant Logic (Prahalad and Bettis, 1986)
Industry recipes in this dissertation could look like:
Enabler: virtual = scalable business model
Actor: large installed fleet of machines = good spare part business
User: good location = structural competitive advantage
Matthyssens et al. (2006) studied five different industries regarding value innovation
and concluded that industry recipes are present in industry networks and managers in
a way that creates inertia for value innovation.
2.2.6 Business models
Markides’ (2006) earlier critique that Disruptive Technology is too technology-driven
and do not take into account business model innovations, is the motivation to explore
business models further. Business models are used to determine how a firm creates a
better value than competitors; thus, it is related to strategy. If successful companies
apply a similar business model, the model can also be related to an industry recipe;
machine and spare parts could be referred to as a razor blade business model, which
has been applied in numerous industries, such as consoles and videogames or printers
Sources of dominant logic
Reinforcement of a world view by market success (operant conditioning)
Complex problem solving behavior
Cognitive simplifications Cognitive bias, available vs. adequate information
Conventional wisdom (paradigms) Past experience and solution by analogy (pattern recognition in chess games)
62
and ink. Business model innovation is not only used for new businesses but also to
reinvent old businesses (Hamel, 1999). Tikkanen et al. (2005) stated that there are four
conceptual levels of management cognition that are related to business models: indus-
try recipe, reputational rankings, boundary beliefs, and product ontologies.
The benefit of business model thinking in systemic changes is that usually business
models consider several aspects of the business simultaneously. Osterwalder’s (2004)
dissertation provided a thorough analysis of different models of strategic thinking, and
he suggested components that should be included in the business model. He also ana-
lyzed the elements of the components, such as fixed pricing (pay-per use, subscription,
or list price). The dissertation also explained that different business models are more
suitable or are at least used in specific industries. Osterwalder later published a book in
collaboration with Pigneur (2010) in which the Business Model Canvas was launched
as a tool, which is illustrated in Figure 9.
Figure 9. Business Model Canvas (Osterwalder and Pigneur, 2010)
At the same time that generic business models, such as razor blade, were in demand,
there was also a call for industry specific models both for existing and emerging busi-
nesses. O’Reilly (2007) published views on the use of the Web 2.0 business model in
the software industry. He began his investigation using fresh examples. His conclu-
sions were not only relevant for Web 2.0 companies but also for firms at large:
Key Part-ners
Key Activi-ties
Key Re-sources
Value Propo-sition
Cost Struc-ture
Revenue Streams
Channels
Customer Rela-tionships
Customer Seg-ments
63
Services, not packaged software, with cost-effective scalability,
Control over unique, hard-to-recreate data sources that get richer as more peo-
ple use them,
Trusting users as co-developers,
Harnessing collective intelligence,
Leveraging the long tail through customer self-service,
Software above the level of a single device
Lightweight user interfaces, development models, and business models
Business model excellence includes the same dilemma as operational excellence, i.e.,
success involves a risk of creating rigidity. Based on their case study, Doz and Koso-
nen (2010) suggested that top management requires three meta-capabilities to achieve
the renewal of a business model of the firm: strategic sensitivity, leadership unity, and
resource fluidity. The Industrial Internet is one concept discussed in this dissertation,
and the way several practitioners define Industrial Internet, takes already position on
some of the elements of the Business Model Canvas.
Digitalization and strategy 2.3
There are a growing number of digital strategy models offered by consulting companies,
an increasing number of digital strategy projects ongoing in firms, and active publishing
of new academic papers about the topic to sustain momentum. Is digitalization a phe-
nomenon that is captured by the analysis of the external environment with the tools
suggested by strategic schools, or can the strategy be developed based on digitaliza-
tion? Does digitalization indeed change the rules of how value is created? The as-
sumption has been that digitalization touches knowledge and value creation in a new
way, which is a potential source of a systemic impact on the economy or on the strate-
gy of a firm. Digitalization also questions what markets can be considered large, like
Chris Anderson is describing in his concept Long Tail12, when Internet due to its global
coverage can transform local niche businesses into sizeable global market opportuni-
ties. The economic principles in the Long Tail are not new but technology has changed
12 http://www.wired.com/2004/10/tail/, retrieved 6.9.2016
64
the application of those principles, which has an impact to business strategies, which
might be also relevant in B2B (Brynjolfsson et al. 2006).
None of the strategic schools recognized the platforms as playing a role in strategy
formation. Open innovation, crowds, and ecosystems are also elements that have been
increasing in scale and impact due to the information network. The concept of a net-
work economy is older than Uber. Platforms and open innovation are reviewed in later
sub-chapters. The link between business models and the Industrial Internet was thor-
oughly examined from different perspectives in the article by Porter and Heppelman
(2014) in which they endorsed the new era of competition due to smart, connected
products and also suggested a path for progress: monitoring, control, optimization, and
autonomy. Porter and Heppelman (2014) also argued that the Five Forces could be
used to understand the industrial dynamics of digital technologies. This normative
checklist was provided for strategists:
Which set of smart, connected product capabilities and features?
How much functionality should be embedded in the product and how much in
the cloud?
Open or closed system?
Full set of smart, connected product capabilities and infrastructure internally or
outsource to vendors and partners?
What data to capture, secure, and analyze to maximize the value of offering?
How to manage ownership and access rights of product data?
Should distribution channels or service networks fully or partially be disinterme-
diated?
Change its business model?
New business by monetizing product data through selling it to outside partners?
Scope expansion?
Digitalization also provides a platform for different concepts for the firm itself. Häcki &
Lighton (2001) argued that orchestration plays a key role in a network economy. They
also provided evidence of network companies that outperformed their non-network
peers. The network economy discussion lost momentum due to an interest in platform
thinking, even though network effects are fundamental for both platforms and open
innovation. Porter and Heppelman (2015) joined the discussion of how digitalization
impacts the operations of a firm when value is created by data. Using analogies from
the software industry, they discussed collaboration (externally and internally) in lieu of
65
hierarchical structures. During transitions, they suggested stand-alone business units,
centers of excellence, or cross-business steering mechanisms. Porter and Heppelman
(2015) also suggested the role of a Chief Data Officer to manage data aggregation and
analytics on the corporate level.
2.3.1 Platform strategy
A forklift can be a platform. A firm that manufactures forklift trucks can choose the core
platform of the truck (e.g., frame, control system, and driver ergonomics) and design
interfaces for partners (e.g., power train, attachment, wheels, and hydraulics). This is a
type of modular thinking that has been common in industrial development for decades.
Platforms have also been important contributors of efficiency for internal operations. In
many car companies, different brands share the same platforms (and components),
and the motivation has been sharing development and production or sourcing costs by
economies of scale. Sawhney (1998) suggested that product platform thinking can also
be expanded to brand, process, globalism, and customers.
Information technology brought platforms to the next level by enabling the development
to be location independent. Also, the core product, software, enabled real-time delivery
economically anywhere. This also expanded the concept of platform products. Can a
platform be built based on niche knowledge, such as knowledge of the wear and tear of
a bearing? Gawer and Cusamo (2002) wrote an interesting book analyzing three firms
that became platform leaders, Intel, Microsoft, and Cisco, and three firms that were
potential platform leaders, Palm, DoCoMo, and Linux. Compared to earlier efficiency
driven modularity, the book described platform strategies that had a symbiotic nature in
which the platform owner and complementors depended on each other, such as a
memory chip and a computer. Their conclusion was that to pursue a platform strategy,
the firm must be able to manage four levers. Scope of the Firm involves the ecosystem,
and Product Technology involves the way interfaces are designed and intellectual
property is managed. External Relationships include the designs of the consensus/
control and collaboration/competition mechanism with the complementors, and Internal
Organization includes the designs of structures, processes, culture, and mindsets when
some of the complementors are competitors. Later, Gawer and Cusamo (2008) offered
two paths for leadership: Coring (if the industry does not have a platform) and Tipping
(winning the platform war through market momentum). The paths have different tech-
nology and business choices.
66
Kenney and Pon (2011) posed a relevant question: if there is an emergent platform in
the industry, which platform is it? Their case study focused on the smartphone industry,
and the potential platform was the OS. The conclusion was that there is a dominant
OS-based design and that Apple leads the platform race; however, due to a conver-
gence phenomenon, a disruptive new logic could emerge, possibly from someone like
Facebook. Like most products, product-based platforms have logical commoditization
risks. Technological products have addressed commoditization both by advancing the
product and/or attaching services to it. Services can also be attached to platforms, but
can service become a platform? Suarez and Cusamo (2009) argued that services can
play a major role in mitigating the commoditization risk of product platforms. Cloud
computing, which is also reviewed later, introduced platforms as a service concept.
Weinhardt et al. (2009) used Salesforce.com as an example of this development. Earli-
er firms invested in proprietary Customer Relation Management (CRM)) systems or
used modules of other systems, such as Enterprise Resource Planning (ERP), for that
purpose. Salesforce.com changed the merging Cloud technology, software, and ser-
vice concept. Zysman et al. (2015) highlighted the essence of algorithms and their rela-
tion to Cloud platforms. They called this change toward service the Algorithm Revolu-
tion.
2.3.2 Open innovation as a paradigm in the development of a firm
Open innovation, or one its forms, crowd, is not only a business paradigm but also re-
flects societal change. Innovation beyond corporate or other borders as a phenomenon
was certainly not invented in recent decades. Western civilization experienced a similar
innovation during the Athens centrum. ICT technology transformed the globe into the
real-time Athens square, and Chesbrough implemented open innovation into manage-
ment discipline, which has been criticized not because of the content but for its novelty,
as suggested by Trott and Hartman (2009). Chesbrough’s (2006) definition of Open
Innovation is:
“The use of purposive inflows and outflows of knowledge to accelerate internal innovation and expand markets for external use of innovation respectively.”
This means that the target can be a firm’s products or services, internal operations
(primary or support processes), or business models. Regarding its relevance to strate-
gy formation, Vanhaverbeke and Cloodt (2014) contended that the overriding sugges-
tion in the strategy literature is to rely on internal resources for development. This
67
statement does not take into account that the tacit knowledge in a value chain has
been transferred as input for internal innovation. Open Innovation also raises the ques-
tion of whether the wisdom of crowds can be superior to (Surowiecki, 2005) the com-
pany’s internal experts.
Open Innovation can be organized by following the traditional principles of subcontrac-
tor relations, whereas the other extreme includes fully open crowd sourcing concepts.
The characteristics of organization can include firm-driven ecosystems, industry trade
organizations, commercial incubators, university organized platforms, etc. Crowdsourc-
ing as a concept for products, services, specific problems, vendors, customers, capital,
ideas, or complete systems solutions is based on the notion that volume, diversity, and
collective intelligence offer additional value. Malone et al. (2010) stated that a collective
intelligence system can be developed by designing it using components that they
called genes. They illustrated the evidence using Wikipedia, Google, and Linux as ex-
amples. Several public buildings are a result of architectural competitions; it was not
called crowdsourcing, but it had the same characteristics. Howe’s (2006) article “The
Rise of Crowdsourcing” suggested crowdsourcing as an alternative for outsourcing.
Competitions take specific forms, such as Hackathons, in which private programmers
spend a weekend developing a solution to a problem provided by an industrial firm.
Hedenborg, the CEO of a Finnish factory automation supplier, Fastems, commented
that their Hackathon provided nine product development initiatives in three different
disciplines.13 The participants were both software students and professional companies.
What drives Open Innovation as a tool for better competiveness? Is it richer content,
cost, speed, or something else? The content is self-evident based on one of the Open
Innovation principles in Chesbrough’s (2003) article, which points out that no one com-
pany possesses all talent, and there are also radical ideas that might be outside of the
field or industry of the firm. The ability to obtain external knowledge might depend on
the autonomy of the people in the experiment. Teece (2007) argued that there is a trap
in Open Innovation if Open Innovation is seen as an alternative to integrating the
knowledge that exists within suppliers, customers, and complementors. He also con-
tended that unique assets and resources are not enough for a sustainable competitive
advantage without unique dynamic capabilities. In his ecosystem framework, he intro-
duced Analytical Systems, which include Learn and Sense, Filter, Shape, and Calibrate
13 http://www.organisaatio-sanomat.fi/hack-the-factory-toi-fastemsille-ideoita-teollisen-internetin-
soveltamiseen/ retrieved 7.7.201
68
opportunities in addition to the external existing and new knowledge. Cohen and Levin-
thal (1990) developed the concept Absorptive Capacity, which they defined as “the
ability of a firm to recognize the value of new, external information, assimilate it, and
apply it to commercial ends is critical to its innovative capabilities.”
Open Innovation can also involve cost issues. In crowdsourcing competitions, the
hours spent on the solutions are not related to the winner’s prize; however, the transac-
tion cost for the incumbents that manage the network of startups or events as well as
the indirect cost of the organizer might be high. Christensen et al. (2005) elaborated on
this in their case study related to technology development and concluded that the rela-
tion and the cost differ depending on the maturity level of the technology. Felin and
Zenger (2014) concluded that the cost related to communication in Open Innovation is
decreasing due to platforms and standardization as well as due to communication
technology development.
Speed translates to cost; unlimited resources can either speed the development or
delay the total process if not managed correctly, as the complexity increases when the
number of people increases in a project. Gupta and Sounder (1998) stated that there is
a growing importance for reduced cycle times in product development; they also
showed higher user involvement in short-cycle product development. Dedehayir and
Mäkinen (2011) suggested that in the systemic industry context, increased technology
based clock speed intensifies time-based competitions in related sub-industries.
Open innovation is structured, expanded, and accelerated by the innomediaries.
Innomediaries can assist in all phases of innovation (Orientation, Exploration, Selection
and Engagement), but their roles vary in different phases (Roijakkers et al. 2014).
Roijakkers et al. (2014) also concluded that some of the innomediaries offer services
based on the expertise of their own personnel, whereas the other type of innomediaries
search and match clients with innovative companies using search engines. Some of
the portals are in the global job market, such as www.freelancer.com, and some focus
on individual disciplines, such as www.design.com.
If Open Innovation is already a new paradigm, what will delay expansion? Obvious
answers are a lack of understanding, knowledge, processes, and attitudes, such as
“not invented here” syndromes. The nature of the barriers most likely differs depending
on the object of Open Innovation. Chesbrough (2010) stated that one of the mental
difficulties in the business model of innovation is that new business models might be in
conflict with the firms’ existing technology. This makes business model innovation more
69
difficult, whether it is internal or uses Open Innovation principles. Huizingh (2011) sug-
gested that Open Innovation expansion has similar characteristics to the diffusion of
innovation, and the followers are reluctant to copy Early Adopters, especially if it re-
quires organizational changes. The slow emergence could also be explained by cultur-
al readiness, as suggested by Savitskaya et al. (2010) in their case study on China.
One major obstacle in Open Innovation expansion is the Intellectual Property Rights
(IPR), whether it is an excuse not to change or a justified concern of valuable intellec-
tual property leaking to current or potential competitors. Economically, it is a question
of whether sharing provides a greater return than keeping the knowledge in-house. It is
a more abstract question than calculating the rate of return for an internal R&D invest-
ment. West and Gallagher (2006) studied the question in an open-source software con-
text and pointed out that sharing innovations with vendors can lead to better products;
they also promoted mechanisms for commercializing the “spillover” innovations. Arora
et al. (2001) asserted that the technology market is growing, which makes the “spillover”
option a real option. As attractive as the intermediate markets for the IPR might be,
they are also more complex to manage than ordinary security markets. Trading IPR
also involves changes in traditional intellectual property management (Chesbrough and
Ghafele, 2014).
2.3.3 The enabling technology behind digitalization
Digitalization in this dissertation is understood as a process in which data is collected,
transmitted, stored, processed, and used for value creation. The input can originate
from people when they translate observations, judgements, or thoughts in writing or
from sensors that collect factual measurement data. In digitization, the information is
coded as a string of ones and zeros in different parts of the process. Coding infor-
mation is common, but ADC (Analog Digital Converter) began the modern expansion.
In the early 1980s, it was used in military applications and was expanded to the civilian
market. Its usage further accelerated in the mid-1990s when ADC became a mass
market component in different consumer applications (Robertson, 2006). There are
numerous digital technologies, but only the fundamental ones are briefly described in
this sub-chapter. They are used in digital systems or concepts that drive the potential
disruption from a technology point of view, which was investigated in this dissertation.
The systemic concepts are described in sub-chapter 2.3.4.
70
A transistor by Bell Labs in 1947 (Early, 2001) was one of the first key components in
the future digital system. It was a simple, small device that could either act as an elec-
tronic switch or as an amplifier. The capacity of the transistor began to increase when
the integrated circuit, which utilized semiconducting silicon, was invented by Texas
Instruments in 1958 (Kilby, 2000). The next step was the microprocessor, which made
it possible to use digital data as input and to process it according to programmable in-
structions. The microprocessor was not invented by Intel, but the Intel 4004 micropro-
cessor from 1971 (Betker et al., 1997) is perceived to be the first commercial micropro-
cessor. It paved the way not only for computers but also for other parts of a digital sys-
tem as an embedded component.
Memory is needed for storing data either temporarily or for long periods. The first hard
discs were invented by the IBM lab in 1956 (Praxley, 2007) to be used as part of their
mainframe computer. Thereafter, there have been physics-based innovations of the
discs or innovations in the production technology, which has enabled the growth in ca-
pacity (measured in bytes, which is equivalent to 8 bits), the decrease in the physical
size of the device, and consequently, the reduced cost of storing data.
Peripherals, such as sensors or various user interfaces, are fundamental for the sys-
tem to interact with the outside world. For entering data or viewing results, there have
been innovations, such the keyboard, the mouse, and the touchscreen. The technical
development of sensors is more important from an industrial perspective. The process
industry measures various factors of the process by sensors in order to control the pro-
cess in a definite way. The importance of the sensors increases when they can be
connected wirelessly to the external systems. This has enabled telemetry, for example,
in which problems with a working machine can be solved remotely before they have
occurred. Systems improve by increasing the amount of sensors that feed the critical
data with a high reliability. The cost also decreases, which further accelerates market
penetration. A robot is an example of a product that could not have been developed
without digital technology. A robot needs sensors for safe navigation and for the input
to make the right decisions, which are made by computer algorithms. Robotics has
catalysed the development of actuators. Work machines require several actuators with
micro movements to operate unmanned. There is also the additional challenge that
both the machine and the load can weigh tens of tons.
Algorithms are the brain of the system. They present the rules of system operation and
are coded with a computer language, known as the software. The history of software is
relatively brief. IEEE software publication place the introduction of the software in late
71
1983 or early 1984 (Ebert, 2008). Thereafter, the development has been rapid, and
there are different generations of computer languages as well as clusters of languages.
Some of the software programmes are platforms, and some are specific to an applica-
tion. Software technology is an interesting topic, but the relevance to this dissertation is
that software is the core enabler of the systemic impact of several digital technologies.
Wireless technologies decoupled various applications based on location constraints.
As disruptive as innovations by Bell, Edison, Hertz, or Marconi were over 100 years
ago in radio technology, more recently, GPS and GSM were the enablers of modern
digitalization. GPS is a technology that utilizes satellites that orbit in space, which origi-
nated in the U.S. Department of Defence. The commercial usage was originally initiat-
ed by the Joint Office as a Navstar GPS program in 1973 (Stotts et al., 2014) that
shared technology with civilian stakeholders. In hindsight, it is easy to conclude that
the program succeeded, as GPS has become a de facto global platform for commercial
applications that utilize location information. GSM technology was developed by enter-
prises and universities that were developing technologies for mobile communication in
Europe. At the same time, there were competing technologies from the U.S. and Ja-
pan. The scale impact occurred because major European countries were able to agree
on the GSM standard (Hillebrand, 2013), which later also became a global standard.
The economics and relevance for GSM and GPS was that they both became global
technologies, and that the industry leaders chose them as platforms to develop applica-
tions for various industries. GSM paved the way not only for voice but also for data
transmission. Data transmission has had rapid developments in speed and cost. Both
technologies were also linked to industrial and consumer businesses.
There are several other technological components in digital systems. The systemic
impact of the individual components begins with the features, but the cost development
has become even more pivotal, creating an increasing number of new applications.
Cost reductions are the result of innovations in production technology and increasing
volumes. The increasing volume originated from the consumer business, but more and
more the same components are also used in industrial applications. The development
has followed Moore’s law:
72
"The number of transistors incorporated in a chip will approximately double every 24 months."14
Moore’s law has been reasonably accurate since the introduction of the Intel 4004.
Development will continue, but the expert’s opinions are not unanimous regarding pre-
cisely how it will follow Moore’s law in the future.
2.3.4 Digitalization as a systemic phenomenon
If the technological components and their features, cost, and impact are relatively fac-
tual, the situation is more complicated for the systemic impact. When purchasing the
next generation of a microprocessor, the capacity, cost, and functions are known. It is
far more complicated to buy one Industrial Internet, even though there are service pro-
viders that claim they can deliver it. Technology based concepts are often transferred
to businesses by the consulting department of technology companies or standalone
consulting companies. The list of digitalization concepts chosen for this dissertation is
based on the rank of the concepts researched and promoted by those companies, and
that are also relevant for a machine-building industry. Other potentially disruptive digi-
talization concepts that emerged during the empirical research are discussed in the
results section. Open Innovation as a potential paradigm changer has been reviewed in
the strategy section, but it is also included in the list. The concepts differ based on their
characteristics, and they can have interdependencies. The scope of the concepts is (in
alphabetical order):
Additive Manufacturing (also referred to as 3D printing)
Big Data/Artificial Intelligence (AI)
Cloud Computing
Internet of Things
Industrial Internet
Model-Based System Engineering
Open Innovation
Robotics (or autonomous operations)
14 http://www.intel.com/content/www/us/en/history/museum-gordon-moore-
law.html?wapkw=moores+law, retrieved 6.8.2015
73
Additive Manufacturing
In manufacturing, there have been numerous technologies, such as machining, weld-
ing, bolting, gluing, casting, and their combinations. The difference in Additive Manu-
facturing (AM) is that the product or component is built by printing it layer by layer,
making it three-dimensional, which is referred to as 3D printing. For AM technology to
advance, the material coverage must be expanded (usually plastic), larger objects must
be enabled, and there must be acceptable speed and unit costs. AM offers systemic
changes; the physical supply chain of products can become digital, or a firm can have
a business model of selling “recipes” for printing spare parts. It is also an enabler of
speedier R&D via rapid prototyping. Engineering a current product to be printed is a
quite different approach from designing a product that was originally designed to be
printed, which is a challenge for engineering competences of the incumbents.
Future development has both product and supply chain paths. By combining the paths
and services, it can catalyse the business model evolution (Cotteleer, 2014). Cohen et
al. (2014) presented five disruptive attributes related to 3D printing: accelerated prod-
uct-development cycles, new manufacturing strategies and footprints, shifting sources
of profit, new capabilities, and disruptive competitors. Gartner’s prediction of spending
on 3D printers is $13.4 billion in the year 2018, and the forecast for 2015 was $1.6 bil-
lion15 D’Aveni (2015) examined annual patents in AM (from 80 to 600 in ten years) and
argued this growth indicating the technology reaching a tipping point. AM is not only
advancing by the high technology companies or system integrators (actors) but also
due to users. The US Navy is printing drones on board vessels for intelligence.16 On
the civilian side, 27 companies in the port of Rotterdam formed an alliance to print four
spare parts after a tedious selection process to be tested in real operations.17 AM is not
only driven by the competitiveness of firms; it also offers potential societal benefits (dis-
ruption for some), such as reducing dependence of current trading partners or delaying
climate change due to a decrease in the unnecessary transportation of products.
15 http://www.gartner.com/newsroom/id/2887417 retrieved at 6.8.2015
16 http://3dprint.com/85654/us-navy-3d-printed-drones/ retrieved at 6.8.2015
17 http://www.maritimeprofessional.com/news/printing-rotterdam-port-274044 retrieved at 6.8.2015
74
Big Data/AI
Big Data involves the economical processing of large quantities of data. It is also
viewed here as one of the steps toward Artificial Intelligence (AI). Interpreting data is
common for humans and growingly also for machines. The change involves the un-
precedented flow of data of human behavior, often internet-based, or data from sen-
sors. At the same time, storing and computing data has become more economical, and
there are better algorithms available.
Big Data has often been referred to as a new source of innovation; the analytics are
somewhat different when making sense of human behaviors or predicting a machine
breakdown. The following two statements are true of machine data: the amount and
nature of the data collected increases rapidly, and data can be transferred wirelessly.
Based on data from sensors, conclusions can often be validated by simulations or em-
pirical experiments based on the laws of physics. The Artificial Intelligence of machines
is related to machine learning or deep machine learning (Itamar et al., 2010). There is
evidence of the effect of Big Data-driven business improvement but also large industry
and firm variety how to capture the benefit (Brown et al., 2014); however, they suggest
that data science requires a center of excellence to develop the required capabilities.
Westerman et al. (2014) emphasized having the customer data in focus when develop-
ing data driven insights. Big Data alone is not the answer for future business success,
as there is always some degree of judgment required, and therefore decoupled conclu-
sions of the company or its culture can lead to strategic mistakes (Guszcza and Rich-
ardson, 2014). Also, the investments in AI will continue, as AI can be the vehicle used
for transforming computing from productivity to content. The Watson IBM is a prime
example in the healthcare business, where advanced analytics assist doctors in diag-
nosing cancer. Likewise, the software giant SAP is developing its HANA technology for
better analytics from predicting football results to business intelligence. Modern tech-
nology can also follow older business models, such as with Google, which has been
using the traditional newspaper business model but more effectively and globally (Hu-
berty, 2015).
McAfee and Brynjolfsson based on their interview of 330 North American executives
concluded that data-driven companies were more profitable (McAfee, Brynjolfsson
2012). They highlighted that the speed and variety of data has increased, which chal-
lenges deep hierarchical decision-making in data-driven environment. McAfee and
Brynjolfsson (2012) came up with five management challenges into the transition to be
75
data-driven: Leadership, Talent management, Technology, Decision making and Com-
pany culture.
Cloud Computing
The dominant process of private and public organizations in developing IT included
following steps until recently: internal process development (as is and to be) and the
choice of suitable software for application and implementation (possible pilots and
rollouts). The cost and lead time was driven by scale, complexity, and the effectiveness
of the process itself (feasibility studies, planning, decision making, and implementa-
tion). Long lead times meant that entities conducted parallel activities (old processes
with the old tools and new processes with the new tools). The Cloud changed this. A
firm can connect to the Cloud and receive a world-class process with IT support in
weeks and can pay for the usage. It is obvious that as a model, it is more flexible than
rollouts and more economical than depreciating the IT infrastructure cost each month.
The US National Institute of Standards and Technology (Mell and Grance, 2011) has
defined three service models for Cloud Computing: IaaS (Infrastructure as a Service),
PaaS (Platform as a Service), and SaaS (Software as a Service), and four deployment
models: Private, Community, Public, and Hybrid Cloud.
From a disruption point of view, the Cloud lowers the barrier for newcomers to enter the
global business market, and it also offers economies of scale to SME’s that were previ-
ously available only for MNEs. Several of the known new business model innovations
are utilizing the Cloud as a technology, such as Netflix in broadcasting. Although the
Cloud is a major source of disruption for the previous type of computing, it is also vul-
nerable to security risks. These issues require both technological solutions and new
policies regionally and internationally (Kushida et al., 2015). Security risks can affect
both public and business domains but also privacy at large. The Cloud has the same
inertia attached to it as AM or AI, as it includes elements of which many incumbents
have no prior experience or knowledge. Iyer and Henderson (2010) listed seven capa-
bilities that are needed in the formulation of a firm’s Cloud strategy: Controlled Inter-
face, Location Independence, Sourcing Independence, Ubiquitous Access, Virtual
Business Environments, Addressability and Traceability, and Rapid Elasticity.
76
Internet of Things
The Internet of Things (IoT) is related to the Internet. Experiments in building computer
networks began in the 1960s in military and university circles. The explosion occurred
in early 1990s with the invention of the World Wide Web. Free access was a prerequi-
site for the explosion according to Sir Timothy John Berners-Lee, who won the first
Millennium Technology Prize in 2004 for his pioneering work with the World Wide Web:
“The decision to make the Web an open system was necessary for it to be uni-versal. Had the technology been proprietary it would probably not have taken off. You can’t propose that something be a universal space and at the same time keep control of it.”18
Thereafter, the Internet has experienced several enhanced technological innovations in
the platform and especially with the applications that utilize the platform. The systemic
impact will occur when people are connected globally at all times. It is easy to under-
stand the magnitude of the value based on Metcalfe’s law from the early 1980s, which
states that “the value of a network grows as the square of the number of its users
grows” (Metcalfe, 2013).
While the Internet is the platform for people to become location-free and have real-time
access to data, IoT involves connecting physical objects. Manyika et al. (2013) defined
it as “the use of sensors, actuators, and data communication technology built into phys-
ical objects that enable those objects to be tracked, coordinated, or controlled across a
data network or the internet.” HCI and M2M were discussed before the Internet, but the
global scalable platform was missing. Some companies, such as Cisco Systems, use
the Internet of Everything for the same concept. Kellmereit and Obodovski (2013)
named five core application domains for the future of IoT: Connected Cities, Homes,
Health, Cars, and Supply Chain. Gubbi et al. (2013) explained that web.3 (ubiquitous
computing) uses intuitive queries and is based on Cloud technology to achieve the po-
tential of IoT. This is also an example of the interdependence of the selected digitaliza-
tion concept, as stated in the introduction of the sub-chapter.
18 Sir Timothy John Berners-Lee, http:/taf.fi/en/millennium-technology-prize/winner-2004/ re-trieved 6.8.2015
77
Industrial Internet
IoT and the Industrial Internet as concepts are mixed both in academic and business
literature and discussions. The core difference is that IoT has a stronger link to con-
sumers, and Industrial Internet is linked to the value creation of a firm, which in many
applications makes it more mission-critical. The ETLA report (Juhanko and Jurvansuu,
2015) presented a model in which the link between a consumer to smart, connected
products and services was lifestyle-based (IoT), for a firm’s data-based business mod-
els (Industrial Internet) and for the society frictionless services (= information society
2.0). GE emphasized the systemic nature by pointing out that the core aim of the In-
dustrial Internet is connecting intelligent machines, advanced analytics, and mobile
people at work (Evans and Annunziata, 2012). The same systemic character is men-
tioned in Porter and Heppelman’s (2014) article in a farming context in which applying
the Industrial Internet concept allows farm equipment manufacturers to move from
equipment performance to farm output optimization. The earlier cited World Economic
Forum (2015) Output Economy concept also described the progress of the Industrial
Internet in which the last three phases represent an increase in the systemic level of
optimization (Figure 5).
In their white paper (Bradley et al., 2013), Cisco estimated the value of the Internet of
Everything (including the Industrial Internet) market as $14.4 trillion by the year 2022,
and the almost equally influential drivers for value creation are: asset utilization, em-
ployee productivity, supply chain and logistics, customer experience, and innovation.
Russo et al. (2014) suggested six questions that should be considered by machine
builders to create value: how does data create new value, how fast will new business
models emerge, how will industry structures change, how do we play in this space,
what capabilities do we require to win, and how do we get started? Individual countries
have also added the Industrial Internet to their agenda to develop or sustain industrial
competiveness. German initiative Industrie 4.0 (Kagermann et al., 2013) is a good ex-
ample, which notably indicates that real-time CPS platforms are changing the paradigm
of industrial service. The US consumer business tradition might explain why the focus
in US industrial initiatives has been from the beginning adding customer value. Also,
the German investment in the Industrial Internet is reflecting the change from the man-
ufacturing to customer in the national Acatech (Kagermann et al. 2014) research report
published a year later than the original Industrie 4.0., which was called the Smart Ser-
vice Welt.
78
Model-Based System Engineering
The focus of this dissertation is machine-builders. Even in standalone situations, ma-
chines are a result of multidisciplinary work. There is an ongoing evolution in the sub-
systems of a machine; material science creates lighter and more wearable structures,
components are becoming more intelligent, and digitalization impacts how machines
are controlled in several ways. The systemic complexity increases when a fleet of ma-
chines or machine interactions with upper systems (planning, maintenance, etc.) are
taken into account. The development of a machine still follows the traditional disciplines
as a foundation (mechanical, electrical, and software engineering), but digitalization
has enabled new sources of input via sensors, new insight by algorithms that utilize Big
Data, and a new set of tools. The tools include both methods such as simulations or
crowdsourcing ideas as well as advancements in software for various disciplines, such
as CAD tools. Model-Based System Engineering (MBSE) has evolved as a promising
discipline to manage the growing complexity of the R&D of a machine builder. It can
combine the physical world with software in an early development phase. Architecture
is needed to enable the actor to create a dialogue between internal and external
knowledge and to orchestrate the use of concepts, such as Open Innovation. MBSE is
an effective method, but further standardization would lead to advancement with influ-
ential actors (INCOSE, 2007) actively using it. Garetti et al. (2012) argued that virtual
prototyping is moving from digital knowledge to a digital lifecycle. This is an important
concept in building new business models that are performance-based. Nichols (2013)
suggested that systemic complex problems can be solved in the CPS environment by a
smart and synchronised development methodology (from previous concurrent devel-
opment), which is also a driver of business transformation. This methodology can in-
clude continuous iteration supported by virtual and real experiments, which is some-
what similar to Scrum (Schwaber and Beedle, 2002) ideology. MBSE is currently in an
early phase, and technologies that can enhance it are still in development.
Robotics
In the context of this dissertation, robotics refers to work machines becoming autono-
mous. All work machines have features that assist the operators or protect the machine
from failures whether caused by the operator, the machine, or the environment. Com-
puters have played a major role in this process. For example, a sensor measures the
temperature of engine coolant, and if the temperature is too high, software instructs the
79
system to shut the engine down before it is damaged. These developments have re-
sulted in the capability to optimize productivity, fuel consumption, tire wear, and other
parameters to improve the lifecycle economics of the machine. The next level of intelli-
gence for machines involves interacting with the environment, including taking instruc-
tions from production planning, communicating with the maintenance system locally or
with a remote help desk, and interacting with the other machines to avoid collisions.
Since machines were invented, operators have occupied the cabin and applied judge-
ment. Replacing human judgement with an algorithm is the essence of robotics in this
research. This is already common with machines in the process industry but not to the
same degree e.g. in farming, mining or port operations. For security reasons, industrial
robotics has separated the man from the machine, but advancements are paving the
way for humans to cooperate with robots in industrial applications.
When a decision-maker’s objective is to switch to automation, it is most likely due to
the anticipation of a better financial return than can be obtained by continuing to invest
in the manual alternative. The technology is relatively novel, and there are concerns
about reliability and safety, which can become better or worse. At the same time, it can
be expected that the underlying technologies are still rapidly progressing. Automation
causes unemployment for machine operators, but it might be more environmentally
friendly; different benefits for different stakeholders. To reduce the inertia in automation
and unmanned operations, that is related to the undesirable societal impact of unem-
ployment; various innovations that enable humans to work with robots should be pro-
moted as suggested by Brynjolfsson and McAfee (2012).
Robotics is also a field in which development in consumer businesses converges with
industrial businesses. There are similar underlying technologies, such as machine sight
or navigation. A self-driving Google car and the service robots in hospitals have more
of an impact on the attitudes of the public than the welding robots in the 1980s. BCC
research has forecasted that autonomous robots make up the largest category of smart
machines (others are expert systems, neurocomputers, smart embedded systems, and
intelligent assistants) with a $13.9 billion annual market.19 Cyber security is a core is-
sue in all digitalization concepts, but autonomous machines are directly vulnerable.
Kohler and Colbert-Taylor (2015) stated that cyber hacking and cyberterrorism are fac-
19 http://www.bccresearch.com/pressroom/ias/global-market-smart-machines-expected-reach-$15.3-billion-2019, 12.8.2015
80
tors accompanied by legal issues that affect the timing of the expansion of autonomous
vehicles.
Summary
Concepts within digitalization are interlinked, and several concepts are systemic in na-
ture. Human behaviors drive the speed of the emergence of digitalization, which is not
only influenced by rational benefits but also by usability and the user interface. There is
not necessarily consensus regarding how the impact of the disruptive nature of individ-
ual concepts differs from industry to industry or for the potential timing. This disserta-
tion has focused on disruptiveness in the machine-building industry, which serves
global ports and terminals. Whether or not views of the value chain differ between ena-
blers, actors, and users was also investigated. Some of the concepts are likely to be
disruptive based on recent advancements and the economics of technology; for other
concepts, technology must continue to advance or become more economical to cause
a disruptive impact. It is also likely that the point of technological advancements or
economics differs between industries. As indicated by the research question, the up-
time is critical for the value chain investigated, which implies that most likely, the way
Uber disrupted the taxi business would not automatically be applicable in mission criti-
cal processes, though there are conceptual leanings from Uber.
Westerman et al. (2012) introduced a matrix in which they defined the digital mastery
level on the vertical axis as dependent on digital capability, and the horizontal axis was
dependent on leadership capability. Applying this definition, they developed four levels:
Beginners, Conservatives, Fashionistas, and Digital Masters. Based on the research
data, they applied the matrix to several industries, which is illustrated in Figure 10. In
this figure, enablers would be closest to high technology industries which are on the
Digital Mastery level. Manufacturing was placed on the Beginner’s level.
81
Figure 10. Maturity by industry, (Westerman et al., 2012)
Synthesis on industry evolution, strategy formation, and 2.4digitalization
The phenomena and mechanisms investigated are part of the industry evolution; the
theory sections discussed seminal theories that explain the dynamics of the industry
evolution, such as the Five Forces (Porter, 1979), the role of the stakeholders (Free-
man, 2015), or emerging theories of industry convergence. The nature of the evolution
was also investigated, which involves stepwise or disruptive continuous improvement.
The focal theories are S-Curves (Foster, 1987), Technological Trajectories (Dosi 1982),
Sustaining and Disruptive Technologies (Christensen 2011), or Diffusion of Innovations
(Rogers, 2003).
As the investigation focused on disruption or paradigm changes, the economics view is
Schumpeterian (2012) in that larger disruptions in economies are linked not only to the
technology or economy but also to politics and society at large. The basic elements of
the factors of production are labor, raw materials, and capital. Knowledge labor is in
key role in digitalization, which has strong links to knowledge-based value creation,
which is illustrated by concepts such as Outcome Economy (World Economic Forum
FASHIONISTAS
Telecom Travel and hospitality DIGITAL MASTERS
High technology
Retail Banking
Insurance
Utilities
CONSERVATIVES Manufacturing
BEGINNERS
Pharmaceuticals Consumer packaged goods
82
2015). Capital and perhaps the more risk-seeking Venture Capital are expected to play
roles in the development.
The catalyst in Schumpeterian (2008) disruption (Creative Destruction) is the entrepre-
neur. In this dissertation, they are represented by the individual firm. A major way for a
firm to cause disruption is related to innovation, which is often referred to as radical
innovation in disruption. The intent, direction, plans, and resources of innovation are
based on strategies of firms. Strategy formation was reviewed based on the 10 schools
of strategies in the classification system developed by Minzberg et al. (1998). The plat-
form is a new element that has emerged in the digital era in economics and strategy
formation. Platforms are also closely linked to Open Innovation and Crowds, which
were included in the discussion of network economy. Five Forces (Porter, 1979) is
very value chain based and Crowds or networks are not conceptually included.
Firms are key players in B2B business. The strategy formation and decisions regarding
resources are made by management. The decisions are influenced by collected beliefs,
such as Industry Recipes (Spender, 1989), Dominant Logic (Prahalad and Bettis, 1986),
and the governance and culture or individual beliefs that have an impact on manage-
ment cognition, which include Bounded Rationality (Simon, 1991), Prospect Theory
(Kahneman and Tversky, 1979), and values. Implementation is likely to be affected by
the same beliefs as well as factors such as leadership style and capabilities, which
include Absorptive Capacity (Cohen and Levinthal, 1990) and Dynamic Capabilities
(Teece, 2007).
Digitalization is a systemic phenomenon, and its ability to enable disruption is likely to
vary based on different theories of economics or strategic management. In addition, the
disruption potential of individual digital concepts varies. Digitalization concepts were
reviewed based on current literature and practices with the assumption that they can
potentially cause disruption within the limits of the research question.
The theory discussion, whether it is linked to economics, industry evolution, strategy
formation, or individual technologies, examined elements that drive progress as well as
elements that inhibit progress. The balance of these forces has an impact on speed
and direction. The forces are easier to analyze individually; however, larger impacts are
most likely due to the systemic impact of the forces as a whole.
The summary of this literature review is illustrated in Figure 11. All potential drivers in
the yellow section include sub-drivers that can work as inhibitors or accelerators. They
83
are based on the characteristics of the driver itself, or they can be caused by societal
and political factors or various beliefs that have been reviewed. The model uses the
Five Forces (Porter, 1979) and Stakeholders (Freeman, 2015) as one theory and con-
vergence as a different type of theory to explain industry dynamics. Radical innovation
represents the means, Digital Business Models, and Networks, Crowds, Platforms as
specific phenomena involved in digitalization. Firm, strategy, and management repre-
sent resources. Capital is specifically linked to Risk Capital. Digital technologies are
AM/3D, Big Data/AI, Cloud, IoT, MBSE, and Robotics tools. Their systemic impact can
be continuous improvement, quantum leaps, or disruption, which might be different in
products, service operations, and business models of an industry.
Figure 11. The research object
Continuous Improve-ment
Quantum leap
Disruption
- +
Sub
-dri
vers
Su
b-d
rive
rs
Su
b-d
rive
rs
84
3 Methodological foundation
Considering the ways digitalization has impacted media and retail businesses, it is fair
to state that a similar disruption or paradigm change has not occurred in the machine-
building industry. This does not exclude the pockets of value creation disruption that
already take place or some of the mechanisms of potential disruption that are similar in
machine-building industry as in industries in which the impacts of digitalization are al-
ready visible. The purpose of the inquiry was to increase understanding; is disruption
likely to happen, what will be disrupted, which factors drive or inhibit the disruption, and
is the view on disruption different in the value chain. The focus of the inquiry includes
innovations whose characteristics and performance are factual, but disruption depends
on how the innovations, tools, or methodologies are applied in the future, which is a
result of the decisions made now by the management or by the choices people do as
employees in the firms or privately as consumers. “Better” technology does not always
win in the market place, which occurred in the Betamax vs. VHS case. The characteris-
tics of the investigation guided the applicable frameworks based on the philosophy of
science, practical methodology, and the research design, which are discussed in this
chapter.
Research philosophy 3.1
Positivism and hermeneutics form the dividing worldviews of science for this disserta-
tion. The positivistic worldview assumes that the findings are true and objective, the
hypotheses are verified as facts or laws, values are not applied, and the research re-
sults are not dependent on the researcher (Guba and Lincoln, 1994). The hermeneutic
85
worldview seeks understanding and posits that the essence of the relation between the
researcher and the environment and the lack of absolute truth does not make the re-
search results unscientific (Gadamer, 2013).
The positivistic worldview of science can address issues of discoveries that enable
digital applications. Magnetoresistive random-access memory, which enabled the
emergence of the hard disc drive, is based on laws of physics; however, Moore’s
law cannot explain the cost or capacity of the memory chips in the future in a way that
would meet the scientific criteria of positivism. As the majority of technologies that can
enable potential disruption are already commercially available, the focus of this re-
search was not to predict the future development of these technologies but to increase
understanding of the impacts on business, which is similar to the ethos of hermeneutics.
The epistemology and ontology of future studies is another perspective in this research.
The fundamental question is whether the future is predetermined or whether there are
multiple futures that are results of today’s decisions. Moreover, does the goal of a de-
sirable future drive these decisions? There are several views among scholars that sup-
port the openness of the future, and concepts such as Futuribles (De Jouvenel, 1967)
and Futures Map (Kuusi et al., 2015) are used. Godet (1994) added action to his con-
cept of Strategic Prospective, which underlines action orientation rather than fatalistic
thinking as a base. Popper (2002) also stated in his harsh criticisms of historicism that
“the future depends on ourselves, and we do not depend on any historical necessity.”
Several possible futures are also the essence of scenario thinking as a methodology.
This does not mean that the future would not be impacted by history or by the current
state of affairs. Van der Heijden (2002) addressed the issue by presenting a diminish-
ing amount of the predetermined issues and a growing share of uncertainty as a func-
tion of time –measured as distance from current affairs. Using this framework, he iden-
tified three ways to predict the future based on distance into the future: forecasts, sce-
narios, and hope.
As the desirable future includes an assumption of value according to Niiniluoto,
knowledge seeking has several commonalities with design sciences (Niiniluoto, 2013).
Based on the use of knowledge, research can be categorized into descriptive and nor-
mative research (Olkkonen, 1994). The objective of normative research is to guide fu-
ture development, and therefore design science can be considered normative accord-
ing to Olkkonen (1994). Malaska (2013) referred to research of the future as futurology,
which includes basic research and philosophy of science questions about researching
86
the future, future studies that involve applying rigorous methodologies, and foresight,
which applies research of the future to making better decisions.
In his book, Creswell (2003) suggested a framework that was used to guide the meth-
odological choices of this research, which is presented in Table 1. Although it simplifies
the schools of the philosophy of science, it offers a current view that considers the di-
lemma between positivism and hermeneutics.
Postpositivism Constructivism
Determination
Reductionism
Empirical observation and measurement
Theory verification
Understanding
Multiple participant meanings
Social and historical construction
Theory generation
Transformative Pragmatism
Political
Power and justice oriented
Collaborative
Change oriented
Consequences of actions
Problem-centered
Pluralistic
Real-world practice oriented
Table 1. Four worldviews (Creswell, 2003)
The research question does not fall into the postpositivism category based on earlier
arguments about positivism. The transformative view has a strong link to political sci-
ence, and therefore is not an ideal guiding framework, either; however, the four charac-
teristics of pragmatism suit the research question. Creswell also pointed out that prag-
matism is not committed to one view in the philosophy of science. Thus, there are ele-
ments of constructivism in this research. The objective was not to identify a framework
or model that explains or predicts disruption or its mechanisms; however, if findings
that can be presented in a descriptive framework emerged during the research, the
characteristics could be viewed as related to the worldview of constructivism.
Digitalization has been described as a set of systemic technologies, as it simultaneous-
ly impacts the business environment, stakeholders of the firm, and the way value is
created. Digitalization can also lead to new types of products, services, operations, and
87
business models. Dussauge et al. (1992) stated that generic technologies can have
numerous applications in various businesses and industries. In a broad sense, digitali-
zation is not a paradigm itself; however, based on the definition of the use of a para-
digm (Kuhn, 1996), as discussed in Chapter 2, there are multiple practical paradigms,
e.g., business models applied by enablers, actors, and users, which are related to digi-
talization. Still, some paradigms involve a broader context, such as whether economies
will be more open or closed. Paradigms can guide scientific work despite the lack of a
standard interpretation according to Kuhn (1996); the same logic is evident when de-
veloping a business. Identifying solid causal truths regarding digitalization as a para-
digm changer in the machine-building industry is unlikely, which creates challenges in
presenting the results of the findings. System thinking offers a viable alternative. For-
rester (1994) stated that system thinking can be applied to system dynamics. In the
same article, he (Forrester, 1994) argued that systems thinking and soft operations
research fit well together in the conceptualization phase of researching system dynam-
ics. He also asserted that the rise of soft operations research partly occurred because it
can address issues of physical and social systems simultaneously.
Mixed methods 3.2
The obvious benefit of mixed methods is the triangulation of qualitative research with
quantitative methods. For this dissertation, qualitative methods were the Delphi inter-
views and case descriptions whereas quantitative methods were used for the survey.
All three methods are discussed in their respective sub-chapters. The validity and reli-
ability of the methods are reviewed in chapter 5.3.
The ethos of this dissertation involves hermeneutics, and there was an intrinsic as-
sumption that the object of inquiry is too complex to achieve a solid understanding;
however, if the mixed methods overlap, are related to the research question, and gen-
erate unique perspectives, triangulation and a holistic understanding can be achieved.
Greene et al. (1989) evaluated the aspects of using mixed method designs and identi-
fied five purposes: Triangulation, Complementary, Development, Initiation, and Expan-
sion. Creswell (2003) classified the research strategy of the mixed methods into Se-
quential, Concurrent, or Transformative. During development, one method feeds other
methods, which could be considered similar to Creswell’s (2003) Sequential category.
Expansion seeks to extend breadth, whereas initiation is related to the discovery of
new perspectives. Greene et al. (1989) analyzed how mixed methods are integrated as
88
part of an analysis or as part of an interpretation. Out of 57 mixed research projects
they evaluated, they concluded that almost half had no integration and that almost 80 %
of those with integration were part of the interpretation.
Using the classification system of Greene et al. (1989) and Creswell (2003), the re-
search design of this dissertation utilized triangulation and expansion for the mixed
method design. The process was partly sequential and partly concurrent so that the
findings from an earlier phase could be used in fine-tuning the following phases. The
sequential and concurrent phases of the research design, and triangulation and expan-
sion are illustrated in Figure 12.
Figure 12. The research design
Delphi method 3.3
The purpose of the investigation was to understand whether there is emerging disrup-
tion enabled by digitalization in a machine-building industry that serves global ports and
terminals as well as the mechanisms that drive or inhibit it. Important assumption is that
dynamics is impacted by differences of the view in the value chain, which has been
simplified to enablers, actors, and users. The source of empirical data was based on
experts’ opinions and their predictions regarding future development. One objective of
the design of the Delphi interviews was a balanced distribution of the experts and their
areas of expertise.
Cases
Survey
Delphi round 1
Delphi round 2
Literature
Research question
89
The Delphi method is not the only option for studying future developments. Other op-
tions include systems and evolution methods, scenarios, other expert methods, and the
creative identification of weak signals (Kuusi et al., 2013). A rigorous system and evolu-
tion analysis would have required a large quantity of data over a long period, which
reduced the feasibility of applying the methodology in a reasonable amount of time.
Scenario research would have been an effective option. It was not selected because it
can be applied later using the results from the Delphi methodology, as Mannermaa
(1999) suggested in his book. Weak signals or important rare events, which Taleb
(2010) referred to as “Black Swans,” could have been used to identify qualitative fac-
tors that can cause disruption. Digitalization is described as a complex systemic phe-
nomenon. Taleb stated that “a complex domain is characterized by the following ele-
ments, both temporal (a variable depends on its past changes), horizontal (variables
depend on one another), and diagonal (variable A depends on the past history of vari-
able B). As a result of this interdependence, mechanisms are subject to positive, rein-
forcing feedback loops, which cause ‘fat tails.’” This dissertation assumes that the ma-
jority of enabling technology already exists and is economical enough to have a disrup-
tive potential, and therefore the focus is on understanding managers’ decisions in ap-
plying the technologies. This also implies that the systemic loops referred to in Taleb’s
definition are more important than identifying new “Black Swans.” There are also expert
methods other than Delphi, such as various workshop methods in which a group of
experts collaborate to solve a problem. The required professional and geographical
aspects of the method would have been both impractical and expensive. Obtaining a
holistic picture implied that some experts would be from competing organizations, and
therefore the anonymity of the Delphi methodology was an integral requirement.
The Delphi method was named after the Oracle of the ancient Greek town of Delphi,
and it got the modern methodology from the Rand Corporation, which investigated po-
tential US industrial bombing targets from the perspective of a Soviet strategic planner
in the 1950s (Dalkey and Helmer, 1963). Linstone and Turoff (2002) characterized the
Delphi method as:
“Structuring a group communication process so that the process is effective in al-lowing a group of individuals, as a whole, to deal with a complex problem.”
One of the reasons to use the Delphi method according to Mannermaa (1999) is to
manage large complex problems between disciplines that lack common experience,
methodology, and language. Rowe and Wright (1999) stated that anonymity, iteration,
controlled feedback, and statistical aggregation of a group response are essential fea-
90
tures of the Delphi technique. Based on the researched cases, they concluded that
Delphi can outperform statistical groups. They also argued that variations in the use of
Delphi methodology increase the difficulty of verifying its validity. Originally, the Delphi
process was intended to create an increasing consensus with a panel by repeating
rounds (Linstone and Turoff, 2002). Kuusi (1999) argued that the panelists’ arguments
of their standpoints (for or against) over a straightforward consensus seeking are more
useful in detecting future developments. Kuusi called this Delphi variant “Argument
Delphi.” There are also special techniques, such as the use of images to encourage
panelists to form an opinion (Adelson and Aroni, 2002). Their study revealed that the
use of patterns created stronger differences in the subgroups. The risk of using images
involves a growing variance in the interpretation of the actual question that might void
the benefit of user-friendliness that using images offers.
The Delphi technique has been criticized. The earliest and most cited criticism was
from Sackman (1974), who based his critique on an extensive review of completed
Delphi studies and concluded that the methodology should not be used until additional
scientific procedures are incorporated. In his list of 16 conclusions, most are related to
the applications of Delphi, such as in selecting experts or designing questionnaires. He
criticized the forced consensus and suggested empirical experiments based on the
methods of social sciences, which implied that the conventional Delphi method would
move closer to positivistic worldview. Kuusi’s (1999) response to Sackman’s critique
was that Sackman criticized Delphi based on opinion polls (transient invariances in
opinion research), whereas the purpose of Delphi should be identifying relevant argu-
ments for future development.
Using modern ICT technology improves the technical applicability of Delphi. Panels can
be organized in which a large number of participants follow the Delphi process and its
features. The development of preprogrammed algorithms enables real-time feedback.
This development favors quantitative methods in analyzing the panelists’ feedback. As
quantitative methods were applied to the survey for this dissertation, face-to-face Del-
phi interviews were conducted to better understand the views of individual panelists.
The source of information for the Delphi method is the expert, which requires the re-
searcher to follow specific guidelines. Experts must be selected in a way that he or she
possesses the necessary knowledge to provide expert answers. The question or the
argument must be formulated in a way that the answer contributes to the actual re-
search problem. The research design must minimize the risk that the expert opinion
reflects personal prejudices, biases, or distortion due to current events or any other
91
beliefs that cannot be completely eliminated. For example, recent digitalization publicity
could cause experts to overestimate the impact, which behavior is described by Availa-
bility Heuristics (Kahneman, 2003).
An important criterion for this research was the richness of data. For the Delphi rounds,
this involved both the roles and backgrounds of the panelists. Kuusi (1999) offered a
framework of the experts’ roles in predicting future developments that can expand the
richness of the data, which is presented in Table 2.
Types of expertise
Scientists Knowledge on invariance
- permanent invariances of non-learning beings
- transient invariances: habits, routines, and equilibrium points of learning actors
Decision-makers Real and perceived capacity limits, perceived interests and routines, and real and perceived capability limits
Synthesizers Relevant invariances, relevant capacity limits, relevant inter-
ests, and relevant capability limits
Table 2. Three types of experts on future developments (Kuusi, 1999)
In Kuusi’s framework, the learning being may have a different behavioral disposition
now than in the future based on experiences; the learning being also has intent and an
active memory. The capacity is based on the resources available to act. According to
Kuusi, real capability limits are boundaries of rational behavior. He also noted that
when real and perceived capability limits are combined, the actor behaves rationally.
Kuusi’s (1999) capability statement is not derived from the competence of an individual
or organization. Kuusi’s General Theory of Consistency posits that a “learning being
continuously changes its cognitive map in response to experiences of inconsistency.”
An important feature is that the decision-makers on a higher hierarchical level can
move beyond the capacity and capability limits of a decision-maker of a lower hierarchy.
A practitioner’s interpretation of the synthesizer is that they master relevance by being
able to operate both in the scientist’s and decision-maker’s space. To illustrate this role,
in the interviews, they are visually placed in the presentation frameworks in between
92
scientists and decision-makers, which is a different visual placement than in Kuusi’s
framework shown in Table 2.
The value chain was a background factor of this investigation. The value chain is a
simplified version of Porter’s (1985) value chain (Figure 7) and it only includes enablers,
actors, and users. Eliminating logistics or any other part of the physical value chain
reflects the focus on value creation that is based on knowledge enabled by digitaliza-
tion. Therefore, the enablers are defined as the firms that supply digital technologies or
related services. Porter’s value chain (1985) is similar to the operations. Operation it-
self can be an object of disruption caused by digitalization. The users are the global
ports and terminals. The framework of experts was designed by merging Porter’s (1985)
simplified value chain with the role of expertise in future development by Kuusi (1999).
As both views consist of three elements, the size of the expert matrix will be 3x3.
3.3.1 Delphi round one
The digital technologies and concepts in economics and strategic management pre-
sented in Figure 11 were derived from the literature. The first objective of the Delphi
interview round one was to determine whether the selection was relevant based on the
judgment of the experts. The validation included all six technological concepts and the
two concepts that are directly related to digitalization (Digital Business Models and
Networks, Crowds and Platforms). The other five concepts were not included in the
validation because they were either seminal theories used to explain industry evolution
or fundamental components in economics, such as firms and capital. The first round of
interviews either added or removed concepts from the pre-chosen 8 (6+2) concepts of
the framework. The likelihood was low, as the concepts are rather generic in nature,
but it is important to mitigate the prejudices that the researcher might have had when
designing the framework based on the literature and research question.
The second objective of the interviews was to identify arguments that might explain
factors that either accelerate or inhibit the concepts of the framework. The framework
included six technological and seven economic/ strategic management concepts. The
accelerators and inhibitors had to be in the same abstraction level in order to be able
draw later on conclusions. The high number of concepts meant that the abstraction
level was higher if the focus would have been one or two concepts. This was mitigated
by encouraging the panelists to express concrete justification or examples when choos-
ing accelerators or inhibitors. The concepts were referred to as drivers, and the accel-
erating and inhibiting factors were referred to as sub-drivers. The interviews were semi-
93
structured. The user interviewees in the first Delphi round were other than ports and
terminals to include the expansion aspect (Greene et al. 1989). The reason for expan-
sion is that there might be information in other industries that has not been learned in
ports and terminals but that is or will be applicable to ports and terminals.
The round one interviews included 2-4 experts in each cell of the 3x3 matrix. For value
chain selection, firms were required to meet the criteria defined in the research ques-
tion, including the expansion in round one. Role section was relatively easy for deci-
sion-makers, as it can be narrowed down to owners, chairmen of the boards, CEOs, or
those who manage a large business (from directors to executive vice presidents). The
same logic was applied for the synthesizer, and the titles typically varied from CIO and
CTO to various titles that indicate that the person is responsible for development relat-
ed to the research question. Scientist selection was more demanding, as the number of
potential persons even from one university was very high. The criteria for interviewees
included publications or experience working with the industry in a specific part of the
value chain that the role the scientist would represent. In all interviews the importance
of the personal expert view was highlighted instead of “official” view of the background
organization.
The process of the structured part of the first Delphi round was as follows:
1. Role in the 3x3 matrix
2. Research question
3. Research design (Figure 12) and research object (Figure 11)
4. Objectives of the interview (drivers and sub-drivers)
An important part of the research object brief was the conceptual distinction between
continuous improvement, quantum leaps, and disruption and the description of the dif-
ferent technologies. The time and style of descriptions was adjusted to match individual
interviews, as all experts chosen possessed high expertise in most concepts, and their
expertise in individual concepts exceeded the expertise of the interviewer. The open
portion of the interview allowed experts to express any perspectives that might be rele-
vant to the research question that were not included in the framework.
Some of the participants were likely to be insiders in public companies, and some of
the questions might be sensitive. Therefore, the interviews were not recorded. The in-
terviewees were also assured that none of the memos would be distributed, no quotes
with names would be used, and no results would be presented in a way that would re-
94
veal their identities. The interviewees would most likely have approved recording, but
they would also most likely have filtered their views. Interesting views of potential dis-
ruption do not necessarily align e.g. with the current agenda of the firm, and recording
was not done even if some details were missed in the written notes. Harvey (2011)
concluded that this is a tradeoff that the researcher must choose for interviews when
trust is of utmost importance. Walsham (2006) argued that recording has a disad-
vantage in that the interview might be less open or truthful. This would increase the risk
of favoring arguments that are better suited to the organization of the respondent. The
outcomes of the interviews included documented views on the disruptiveness of pre-
sented drivers and suggested sub-drivers.
3.3.2 Delphi round two
The objective of the second interview round was to rank the selected drivers, and the
three most important sub-drivers that either accelerate or inhibit drivers. The suggested
sub-drivers were derived from the literature and Delphi round one. The experts were
permitted to present new sub-drivers as well. When a new sub-driver emerged, that
sub-driver was available for the next interview. The pattern of the emergence of new
sub-drivers and their order of rank provided additional data on the consensus patterns
in the value chain. Delphi round two also included an open portion in which the experts
could either elaborate on their rankings and suggestions of new sub-drivers or express
their views of the research question regarding whether digitalization is disruptive for the
machine-building industry, which part of value creation is mostly impacted, or which
mechanisms cause disruption.
The selection of the population was done based on the same principles as described in
Delphi round one with two exceptions. The size of the population was expanded to four
experts in each of the cells to enrich both international coverage as well as the cover-
age of the firms and universities. Some of the experts were the same as in round one.
The second difference was that the users were only from ports and terminals. For uni-
versities, this implies that the experts had researched transportation or logistics. This
increased the focus on the research question. The interviews were not recorded based
on the same reasons as the round one Delphi interviews.
For the structured questions, either VAS (Visual Analogue Scales) or Likert scales
could have been used. The motivation to use order ranking was to encourage experts
to think critically. The use of a Likert scale could generate average ratings, which would
yield similar values for several of the drivers and sub-drivers. This result could have
95
been avoided by eliminating the midpoint of the Likert scale, which according to Gar-
land (1991) is effective for socially unacceptable answers. Fabbris (2013) showed that
the mental effort further increases during order ranking than rating, which suggests that
for order ranking, the choices should be below 10. The procedure can be simplified by
concentrating first on the highest and lowest. For the drivers, this was achieved by
ranking only six digital technologies and seven economic/ strategic management driv-
ers. The selection of only the 3+3 most important drivers for the sub-driver ranking was
partly done to simplify the mental effort, as Fabbris (2013) argued. The second reason
was to maximize the interview time so that issues that the expert judged to be most
relevant to the research question could also be discussed.
The experts were told that the logic of disruption is based on the Schumpeterian (2012)
concept, which was reduced to the PEST model from which drivers were derived from
digitalization (technology in PEST), economic/strategic management (Economy in
PEST) literature, and Delphi round one interviews. Political and social factors were
present (if deemed applicable) in the sub-drivers that either accelerated or inhibited the
digital or economic/ strategic management drivers.
The process for the structured part of Delphi round two was as follows:
1. Briefing
2. Order ranking of technology drivers
3. Order ranking of economic/strategic management drivers
4. Technology (digitalization concepts) driver number one
a. Suggestion of new inhibiting sub-drivers
b. Order ranking (1,2,3) of the inhibiting sub-drivers
c. Suggestion of new acceleration sub-drivers
d. Order ranking (1,2,3) of the accelerating sub-drivers
5. Economic / strategic management driver number one
a. Suggestion of new inhibiting sub-drivers
b. Order ranking (1,2,3) of the inhibiting sub-drivers
c. Suggestion of new acceleration sub-drivers
d. Order ranking (1,2,3) of the accelerating sub-drivers
6. Repeating the same process as no. 2 and 3 technology and economic /strategic
management drivers
7. Free comments
96
The presentation slides used for the panelists are shown in Appendix 2.
Survey 3.4
The first objective of the survey was to triangulate the results from the Delphi interviews
for all digitalization drivers and economic/ strategic management drivers that are direct-
ly linked to digitalization (Digital Business Models and Networks, Crowds, Platforms).
As the respondent’s time spent on the survey was considerably less than the time
spent by the interviewees, Digital Business Models were simplified to the Industrial
Internet, which is assumed to be better known, and Networks, Crowds, Platforms were
simplified to Open Innovation for the same reason. To enable triangulation, participa-
tion followed the same 3x3 matrix as the Delphi interviews. The objective was to
achieve a minimum of 250 responses.
The survey also expanded the breadth and scope of the inquiry (Greene et al. 1989) of
the Delphi interviews in three different areas. First, there was an expansion to the
judgment of the explicit systemic impact of the digital technologies to firms’ products,
services, operations, and business models. The second expansion was the participants’
judgments on timing if they perceived that the systemic impact of the digital technolo-
gies would be disruptive. The third expansion involved understanding how well the
firms or research institutes are prepared for the potential disruption based on the re-
spondents views on future development. This implies that a small investment may be
adequate if the digitalization impact is perceived to be low.
The reason for the targeted size of the population is that it enabled conducting statisti-
cal tests using the background factors (independent variables), which in this case are
the respondents’ place in the value chain and the role of expertise in predictions of the
future. A third background factor could be the respondents’ judgments on how well their
organization is prepared for potential digitally enabled disruption. Values assigned to
the various technologies and concepts, and the impact to products, services, opera-
tions, and business models also offer the possibility to analyze dependencies using
statistical methods.
3.4.1 Experiment
One issue related to surveys is a low response rate according to a study by Baruch
(1999) in which data from 175 different studies was used. His result showed an aver-
97
age response rate of 55.6 %, but only 36.1 % in studies in which respondents were
from top management or companies. For surveys, there are also unknown factors, in-
cluding who actually completed the questionnaire and whether or not the respondents
understood the questions in the same way. The design of this survey attempted to miti-
gate these problems using a method in which the respondents were recruited from
seminars, either organized by conference companies, research institutes, or compa-
nies internally. Seminars were selected so that almost all audience members would
represent one of the roles in the matrix. The survey began by presenting the definition
of the concepts used in the survey, explaining the response process, and instructions
regarding how to complete the expert matrix position. The objective was diversity in
both companies and geography. This process design provided close to a 100 % re-
sponse rate and fewer misinterpretations of the questions than in mailed question-
naires. One risk is related to the question of whether representatives in the seminars
were from a relevant population. The study by Short et al. (2002) revealed that based
on 437 studies, less than 20 % used random sampling. They also evaluated one study
related to CEO performance and concluded that different sample strategies provided
different results. This survey included a hybrid approach in which company internal
seminars typically have a predetermined population, but open seminars have a more
random audience. The obvious risk is that the nine cells could be either underrepre-
sented or unevenly distributed. The matrix offers a possibility for comparison by aggre-
gating the value chain roles or doing the same for the expert roles, which would both
provide a group of three for comparison. The whole population could also be used as a
statistical comparison group against the aggregated Delphi interview results.
A mixture of a structured scale and a VAS scale was used as a measurement tool,
which was applied so that respondents first determined whether the impact of an indi-
vidual question was continuous improvement, a quantum leap, or disruption. Under the
chosen category, the respondents indicated the strength of the impact by placing an “x”
on a continuous scale in which the furthest left is the weakest impact, and the furthest
right is the strongest. During the analysis, the three categories were then divided into
three sections so that continuous improvement in the first section received a value of 1
and last section received a disruption value of 9. The categories were marked with col-
ors to differentiate between them, but there were no red, green, or yellow colors to
avoid any impression of desirability. The Likert scale and its derivatives are predomi-
nantly used in business research, but VAS is most effective when a respondent has a
strong view on a topic but some of the reasons might be difficult to explain. Based on
the definition of VAS in research articles, it is mainly used in medical research, such as
98
when patients must rate the degree of pain. The expansion of VAS to other sciences
seems to be related to online surveys, which can partly be explained because the use
of VAS with pen and paper has an extra phase due to recording the VAS result into a
numerical value. In online surveys, this can be programmed as an automatic feature,
and there are VAS engines available on the Internet. Neibecker (1984) studied struc-
tured scales and continuous scales with two test groups and concluded that there were
no major differences in the results; however, he determined that the continuous scales
might have better user interfaces, as they are less monotonous and also allow for more
accurate answers than structured scales. The VAS results from this research were re-
structured; the raw data remains available for further research using also the detailed
values.
The experiment proceeded in the following sequence:
1. Background (research question, definitions, process, instructions for the role)
2. Answering section one (impact to products, services, operations, business
models) individually with a 30 second time limit for each judgment
3. After section two, the participants indicated disruption in any of the categories
were instructed to have a view how long it takes that the disruption has reached
a status that it is a normal way of doing business, 60 seconds
4. In section three, the instructions were repeated, and thereafter, participants in-
dicated their views on the disruptiveness of individual technology and econom-
ic/ strategic management drivers (120 seconds).
5. Finally, the participants rated the readiness of their organization in 30 seconds.
The data collection was formatted with exact timing. The same presentation slides
were used, and each slide had the same script to minimize variance between the ex-
periments. The total duration of the experiment was 20 minutes. The survey form is
shown in Appendix 3.
3.4.2 Analysis
The survey data was filtered and completed before the analysis. During filtering, sam-
ples that were missing any background information (independent variables) were ex-
cluded. The final question was answered using the VAS scale (which had guiding
numbers shown in Appendix 3) to indicate his or her organization’s preparedness for
99
the potential disruption, which is an interesting result, per se; however, the answer was
also used as additional background information. This was accomplished by dividing the
VAS scale in three equal sections so that the answers close to 0 were neutral, answers
closer to -2 were regressive, and the answers closer to +2 were progressive.
The pen and paper method was used for the survey, and after filtering, the data was
typed into the Excel database so that each row represented one respondent. After the
initial input, all data points in the database were checked twice to minimize marking or
typing errors. Completion was done by replacing the missing value in the sample with a
value that was the average value of the variable given by respondents in the same
role/value chain matrix. Completion reduces variance, and it is selected to be applied
based on the total amount of missing values in the whole population.
After completion, whether the data followed a normal distribution was verified, as this
impacts the choices of analysis. This was done using the Shapiro-Wilk test (Shapiro
and Wilk, 1965), which according to Razali and Wah (2011) is the most powerful of the
four normality tests they evaluated as long as the population size is above 30. The
Shapiro-Wilk test as well as other analyses and visualizations were completed using
the Microsoft Excel add-on statistical tool XLSTAT20.
The role of expertise in future (Kuusi, 1999) creates different perspectives. The object
of inquiry was to understand how the views differ based on where the respondent is in
the value chain, and the role of expertise is used to enrich the perspective. Also, the
additional independent variable, the preparedness of respondents’ organizations for the
potential disruption, might explain the given values. This creates a risk of generalization
of the results if the size of sub-populations differs, as some of the sub-populations can
either dominate or be under-represented. For generalization, the size of all populations
should be between 10 and 20 (tight control); otherwise, it should exceed 30 (Roscoe,
1975). Forcing equal representation later on would create a risk that the new samples
would not be comparable with old samples. The generalization risk can be mitigated by
balancing the sub-populations in which each sample is assigned a coefficient that bal-
ances the contribution of each sub-population in the entire population. To obtain the
coefficient for individual samples, the total weight of the sub-population is divided by
the number of samples in the corresponding sub-population. In further analyses, se-
lected methods can utilize the coefficients when calculating weighted averages or vari-
20 https://www.xlstat.com/en/solutions/base
100
ances, for example. The data formed 27 sub-populations when using three independ-
ent variables (3x3x3).
The final step before presenting the results was to ensure that the interaction effect did
not distort the interpretation of the results. The interaction effect (Dawson, 2014) shad-
ows the true impact of the background factor as another background factor (also called
a moderator) that affects the same value of the inquiry. The background factors of this
inquiry were role, the place in the value chain, and perceived investment impact. The
presence of the interaction effect in the balanced data was determined by running a
two-way ANOVA test.
The survey results are initially reported as individual results. The interpretation of the
results is included in the discussion and conclusions Chapter 5 in which the results are
compared to both the literature and the results using the other two methodologies.
Cases 3.5
The objective of the case descriptions was to illustrate that there are existing or emerg-
ing disruptive phenomena in the scope of the research. The descriptions indicate future
developments through analogies, and they stem from De Jouvenel’s (1967) claim that
the future must be plausible based on the current state of affairs. Analogies do have
shortcomings, which Martino (1993) listed as: lack of inherent necessity (history was
not based only on physical factors), historical uniqueness, historically conditioned
awareness (do your homework well), and causal analogy (things are not normally the
same in all respects). He also emphasized the need for holism when using analogies,
as technological change is a result of technological, economic, managerial, political,
social, cultural, intellectual, religious-ethical, and ecological dimensions, which is a
similar viewpoint as PEST (Aguilar, 1967) and its derivatives regarding changes in the
business environment of a firm. In this dissertation, analogies serve a hermeneutic
purpose to increase understanding rather than to predict causality.
Case research is a widely used method in management science, and case methodolo-
gy has been an integral part of the research of multiple highly cited articles and books
(e.g., Chandler 1990, Christensen 2011, Gawer, and Cusamo 2002). The number of
cases needed is based on the goal of the inquiry. A study by Benbasat et al. (1997) on
the applicability of case research in information systems showed that if the aim is de-
101
scriptive, a multiple case approach is recommended. Yin (2014) listed sources for case
research data: documentation, archival records, interviews, direct observations, partici-
pant observation, and physical artifacts. Case studies have similar sampling issues as
other methodologies, but the methodology is chosen based on the aim of the investiga-
tion. Eisenhardt and Graebner (1997) discussed sampling in the theory building that
utilizes case research:
“The purpose of the research is to develop theory, not testing it, and theoretical (not random or stratified) sampling is appropriate. Theoretical sampling simply means that cases are selected because they particularly suitable for illuminating and extending relationships and logic among constructs.”
The sampling in this study followed the view of Eisenhardt and Graebner (1997), espe-
cially because the role of the case descriptions involved triangulation and expansion.
The criterion for the cases was that the firm was either conducting or planning (pub-
lished) to conduct business utilizing digital technology to create value in a novel way.
Value is also created in a space between the current actors and users, and the case
actor is not a direct competitor of the digital enabler of the case. The value potential
must be considerable in comparison to the investment needed. In the case descriptions,
the structure of the industry is described as well as how the case has created new val-
ue and how it is away from actors (either from the current business or future growth
potential). The case data was collected from public sources but validated by the repre-
sentatives of the case firms or by the representative of actors impacted by the cases.
The detailed case design was completed after reviewing Delphi one round and the pre-
liminary results from the survey, which follows the sequential approach of mixed meth-
ods (Creswell, 2003).
The selected cases include Enevo21, Trimble22, and XVELA23. Enevo is a firm involved
in waste collection, Trimble is involved in farming (the business case has several simi-
larities with the case presented by Porter and Heppelman (2014)), and XVELA is in-
volved in container shipping and handling. Detailed descriptions will be presented in
Chapter 4.
21 https://www.enevo.com/
22 http://www.trimble.com/
23 https://www.xvela.com/
102
4 Results
The results are presented by following the same mixed methods logic as in the previ-
ous chapter, and they form stand-alone sections. The comparison of the results by dif-
ferent methods is in the discussion and conclusion (Chapter 5), which is the inherent
strength of the mixed methodology research design. The user in this enquiry has been
a global terminal operator. Therefore, the first sub-chapter, 4.1., presents the charac-
teristics of the global container handling industry in order to create a background
against which the results can be understood.
Global container handling industry 4.1
The generic industry is transportation, which has a sub-segment—cargo transporta-
tion—to separate it from passenger transportation. The focal mode of transport is sea
transportation, even though land transportation is briefly discussed as part of the sup-
ply chain. Over 80% of global trade is seaborne (UNCTAD, 2015). The alternative
mode to long-haul ocean shipping is air freight, whereas the short-haul shipping alter-
natives are transport by road, rail, or river. In sea transportation, the transported goods
are liquids, gas, bulk, and break bulk. The containers are steel boxes used to carry the
cargo, and are standardized in length, width, and height; the most common size is 40
feet long, 8 feet wide, and 9.5 feet high, the volume of which determines the maximum
dimension of the goods to be loaded into the container. The majority of the container-
ized cargo is break bulk, although even bulk and liquids can be transported in contain-
ers. A reefer container is a special container with an integrated cooling unit to keep
perishable goods (e.g., food) in ideal condition during the voyage.
103
The transportation system has two main actors: the shipper and the consignee. The
shipper or consignee can be a primary producer, factory, trading company or individual
consumer, depending on the function of the transported goods.
Ocean shipping originated hundreds of years ago, when the sailing ships were carrying
spices from Asia to Europe. The loading and unloading of ships was manual work.
Based on this dissertation, one can argue that the invention of the container in 1956 on
the East Coast of the United States (Levinson, 2006) was a disruptive innovation in sea
transportation. One of the interviewees in the first Delphi round commented that disrup-
tive innovations rarely satisfy fundamentally novel needs but rather satisfy old needs
with a fraction of the cost. The container eventually met the criteria. Using today’s lan-
guage, one could say that the container was a platform that enabled disruption of the
transportation system. With this view, the network effects were important—the number
of actors who adapted to the new method and the degree to which freight was contain-
erized. One important characteristic was that the container set the international stand-
ards that complementors were able to use in their development—ship, truck, rail wagon
and container handling machine builders. The container was a very important decision
to a port, as the weight of the container—tens of tons—made it impossible to handle
with manual labor, leading to the need for mechanization of ports and, consequently,
the pioneering innovations in container handling equipment. The end result of the inno-
vations made transportation costs of many goods so low that the goods could be pro-
duced wherever it was most economical, regardless of the final place of usage. The
innovations also created inertia, where the cost of changing the container as the plat-
form of the transportation system had become beyond justification.
The user in this dissertation is a port operator in the service business. The needed in-
frastructure (e.g., quay, buildings, power supply) is often leased from the port authority.
The main supplier for the port operator is the equipment manufacturer which provides
the equipment needed for the operation. For running the port, the operator needs an
array of services like ICT services, building and equipment maintenance, and security.
The operator can opt for most of the services to be provided internally or by external
vendors. The container operator—one of the port stakeholders—serves the shipping
lines in the sea transport system and the trucking and railroad companies in the land
transport system. If the operation is transshipment, the containers normally arrive in
large vessels and are redistributed to continue transport by smaller vessels, and vice
versa. Customs, freight forwarders, and labor unions are other typical port-specific
stakeholders.
104
Industry has always been international due to the nature of shipping. The shippers and
consignees are extremely fragmented, but the shipping lines, ports, and equipment
suppliers have been consolidating over the years. The five largest shipping lines ac-
count for 48.7% of the global market (Alphaliner, 2016), while the five largest terminal
operators represent 49.7% of the market (Drewry Maritime Research, 2014). In addi-
tion, the five largest equipment suppliers provide 87% of industry cranes and 66% of
mobile equipment; three companies were present in both categories (Schäfer, 2014).
Some of the shipping lines also operate terminals, which typically are also open for the
other shipping lines. The five largest shipping lines appear in Table 3, and the five larg-
est terminal operators in Table 4. The container flows are driven by consumer markets,
and the highest growth occurs when there is a growing middle class. For example, in
2013, 54% of the world’s containers were handled by Asian ports (Drewry, 2014).
Rank Operator Existing capacity /
million TEU
Market share
1 APM-Maersk 3.1 14,7 %
2 MSC 2.7 13.0 %
3 CMA-CGM 1.8 8.8 %
4 COSCON 1.5 7.4 %
5 Hapag-Lloyd 0.9 4.5 %
Table 3. Five largest shipping lines (Alphaliner, 2016)
Rank
2013
Operator Throughput / mil-
lion TEU
Market share
1 Hutchison Port Holdings 76,1 11.8 %
2 APM Terminals 68.0 10.6 %
3 PSA International 61.7 9.6 %
4 Cosco Group 59.9 9.3 %
5 DP World 53.7 8.4 %
Table 4. Five largest terminal operators (Drewry Maritime Research, 2014)
105
The business environment for the shipping lines, ports, and equipment manufacturers
can also be analyzed using the same framework, PEST (Aguilar, 1967), which was
utilized in the Delphi interviews. Political connections to the maritime industry are
strong. Merchant shipping, and an own navy to protect the sea trade have strong roots.
Functioning ports have equally key functions in the foreign trade of a nation. Historically,
ports were run either by governments or municipal organizations, but one strong trend
over recent decades has been privatization; the global terminal operators have taken
the operational role. The largest operator, Hutchison, is running terminals in 52 ports
located in 26 countries across all continents.24 However, the public authorities have
kept their control through the terms of the concessions—often over 20 years—that they
have granted to operators running their terminals. The concerns about climate change,
environment, and safety are embedded in the political goals that end up in the legisla-
tion and regulation that impacts the maritime industry. Ships and the port equipment
use fossil fuels, and carbon directives, engine emission regulations, and fuel taxation
are measures that politicians can use to mitigate the impact. The rising cost of
noncompliance, like oil leakage, also drives behavior to use technologies that are less
oil-dependent. Environmental permits for greenfield ports call for stricter procedures
than those of the existing ports. Ships, shipyards, and ports are also hazardous
workplaces, and various types of safety standards are common. Ports are also the
places where the goods not listed in the manifest are flowing (e.g., drugs, arms, explo-
sives, radioactive and chemical materials, illegal human immigrants), which have both
direct and indirect impacts on maritime stakeholders.
The economic impact of the ocean shipping and port handling costs are relatively low
in relation to consumer goods transported in the containers. Therefore, the efficiency is
not driven by fragmented shippers or consignees; rather, the cost is levied into the re-
tail prices. However, shipping is a volatile business, and the shipping lines have large
investments in capacity. The competition has kept earnings at a relatively low level,
and the EBIT margin has varied between 5-10% between 2006 and 2013 for the major
companies (Schäfer, 2014). Shipping lines have mitigated this volatility by forming alli-
ances to, for example, share capacity in ships. In 2012, the size of the container liner
business was $225 billion (Schäfer, 2014). The port business is location-driven; if a
port is near a large consumer base, it already has a structural advantage. The opera-
tional success factor for the port operator is to turn the ship fast and economically, as
24 https://www.hph.com/en/webpg-87.html, retrieved 13.1.2016
106
the ship is the largest investment in the system; 18.000 TEU ship cost approximately
$185 million25. It is equally important that the turnaround time is predictable, which al-
lows the ship to travel at an optimal speed in order to save fuel. The revenue for the
operator comes from the volume of containers multiplied by the handling fee per con-
tainer; the global market size was $57 billion in 2012 (Schäfer, 2014). The four main
elements in the cost side for manual operation are capital cost of the equipment, driv-
ers, fuel, and maintenance and repair. Port operators are making far better profits than
the shipping lines; the average EBIT margin varied between 20-30% for the major
players between 2006 and 2013 (Schäfer, 2014). Equipment suppliers can be catego-
rized into regional or global; they can also be divided into general machine builders that
also make port equipment, or specialized companies. The business model is building
the equipment fleet in a competitive market in order to make the profit with spare parts.
Various levels of service contracts slowly emerge. Distribution is a hybrid of direct sales
and use of distributors. The size of the port equipment market was $5.7 billion in 2013
(an increase of 10-15% after service revenue), whereas the EBIT margin has varied
between 2-6% with major players between 2007 and 2013 (Schäfer, 2014).
The strongest social connection is employment. Indirectly, a functioning port supports
employment through foreign trade in the same way that ineffective ports drive employ-
ment away. This critical role has made port workers’ labor unions strong throughout the
world. The maritime cluster—shipping lines, shipyards, ports, port service providers,
port equipment manufacturers, subcontractors—is also a large source of direct em-
ployment. Importance for employment has made governmental subsidies in shipbuild-
ing, for example, a common phenomenon. In Finland, the maritime cluster employed
47,000 people (Viitanen et al., 2003).
Port-related businesses encompass multiple technologies, but only the ones that im-
pact the system are mentioned (i.e., ship size, electrification, automation, smarter ma-
chines, and ICT). The ship size is undoubtedly the largest driver of the dynamics, as it
impacts the economy of the business, freight routes, seaside operations, ports, and the
hinterland, and it drives both the size of the container handling equipment—particularly
cranes—and all other technologies that can speed up the operation. Figure 13 shows
the development of container ship size in the last 20 years.
25 http://www.bloomberg.com/bw/photos/2013-09-05/holy-ship#slide2, retrieved 13.1.2016
107
Figure 13. Ship size development in TEU's (Drewry Maritime Research, 2014)
Electrification is one technology trend that relates to the endeavor to reduce fossil fuels.
It makes the local emissions insignificant, but the systemic impact is dependent on how
the primary electricity is produced (e.g., oil, gas, coal, solar, wind). Electricity also of-
fers potential to harvest the waste energy when lowering a container or braking mobile
equipment, for example, when the waste energy can be stored into the batteries, simi-
lar to the process in hybrid passenger cars. Passenger cars, in fact, are the primary
industry where heavier batteries are currently researched and developed. Local emis-
sions are also important, as the ports are close to urban areas. Technologies that ena-
ble automation are driven in some parts of the world by labor shortages, in other areas
by the cost of labor, and in some applications by safety; these drivers also exist at the
ports. The first unmanned terminal was opened in Port of Rotterdam by European Con-
tainer Terminals (ECT) in 199326. Since 1993, the number of automated terminals in-
creased from one to 26 in 2013, and at that time there were plans for an additional 30
unmanned terminals, which indicates that the market is breaking (Schäfer, 2014). The
ship loading and unloading is still done by the operators, however. Smarter machines
are present in the ports, but the situation is no different than in mining, construction,
agriculture, or forestry sectors. The machine components have embedded software
and sensors, and the machines can harvest the data for the purpose of internal control
and optimization; they can also communicate externally, either locally or over the Inter-
26 http://www.ect.nl/en/content/history, retrieved 18.1.2016
5,000 0 10,000 15,000 20,000
1995
1999
2003
2007
2011
2014
108
net. ICT technology is embedded in components, machines, fleets, and system levels,
as well as within the processes of the firms that build the machines.
It is important to recognize that shipping has always been, and is likely to remain, fun-
damentally a commodity business (Lorange, 2009). This implies that despite the choic-
es of specialized strategy, there is a prerequisite of cost competitiveness in products,
services, and internal operations for the actors participating in shipping. It is fair to as-
sume that many of the old recipes of continuous improvement still work. Digitalization is
also a current topic in shipping, and the recent McKinsey and Company research report
argued that digitalization is not only enabling better, safer, and simpler processes in
shipping, but also reducing significantly costs in all phases (Hausmann et al., 2015).
Results from the Delphi interviews 4.2
The first Delphi interview round took place between October 20, 2014 and February 12,
2015 (comprising 92% of the interviews), and the second round interviews were con-
ducted between October 19, 2015 and December 21, 2015 (comprising 95% of the
interviews). The combined number of interviews was 60, from which 54 were face-to-
face; and six interviews were online. Interviews lasted between one and three hours,
with more than 80% lasting around two hours (the time predicted when agreeing to the
interviews). The two-hour time was based on three test interviews conducted prior to
the actual interview rounds. The list of participants of both Delphi rounds one and two
are located in Appendix 4, and after each panelist there is a reference in which value
chain perspective he or she was interviewed (E= Enabler, A=Actor and U=User). This
was important as some of the panelists had been working during their career in more
than one value chain area.
A general observation from the interviews was that expert behavior differed depending
on the role of expertise on the future (Kuusi, 1999). Decision-maker interviews were
clearly shorter than those of the scientists. Decision-makers expressed quite often the
logic of uncertainty, but at the same time the need to make decisions based on what is
likely to happen despite imperfect information. Scientists, independent of their location
in the value chain, took longer than decision-makers to form their views on specific
drivers or sub-drivers, despite the fact that their actual knowledge often was deeper
than the knowledge of their decision-maker counterparts for that specific driver. Most of
the scientists, however, offered valuable input to the research methodology even when
109
it was not asked as part of the interviews. The synthesizer group was closer to the de-
cision-makers in their ways of interacting in the interviews. This most likely can be ex-
plained by the fact that they were employed by the businesses, even though their func-
tional responsibilities were close to research and development.
4.2.1 First Delphi interview round
First round participation based on the expert role in the value chain is illustrated in Ta-
ble 5. Panelists came from seven different universities or research institutes, and 14
different companies, of which 12 were in global business. One-third of participants
were not Finnish citizens.
Role/ Value Chain Enablers Actors Users
Scientists 3 3 2
Synthesizers 4 3 2
Decision-makers 3 2 2
Table 5. The Delphi panel in round one
The first objective of Delphi interview round one was validating the six technological
drivers that that were suggested by the earlier cited research reports (like Manyika et
al., 2013) to have potential to cause industrial disruption. The suggested drivers were
AM/ 3D printing, Big Data/AI, Cloud, IoT, MBSE, and Robotics. Validation included a
possibility to suggest new drivers. The second objective was finding sub-drivers that
either accelerate or inhibit the development of each driver.
The validation was an open discussion, where each driver was reviewed in the context
of continuous improvement, quantum leap, or disruption. Review was not based on
scoring or ranking but letting the panelist elaborate what his or her judgment of the im-
pact of a particular technology. The researcher translated later the written views into
the above mentioned three categories. A comment like, “I believe that 3D printing will
change the rules how manufacturing companies create value” implied disruption cate-
gory.
110
Out of the 24 interviews, 57 views were given that the driver could be disruptive in the
context of the research question, 41 that the maximum that the driver could reach is a
quantum leap in competiveness, and only five that viewed the driver as unable to
achieve more than continuous improvement. The span for the individual drivers ex-
tended from two to 16 views being disruptive. The driver that got only two disruptive
views got 14 views being a source of a quantum leap. Based on the results all original
drivers were included in the order ranking process in the Delphi round two. One panel-
ist proposed mobility as a new technological driver with a context of firms’ resources
being location free. It was not taken to Delphi round two on the basis that mobility is
inherent feature of the Networks, Crowds, Platforms which is one of the economic /
strategic management drivers.
Sub-drivers were also derived from the written comments where the panelist elaborat-
ed his or her judgment of a particular technology. A comment like “3D printing will be
disruptive technology as soon as metal based materials can be printed” implied the
inhibitive sub-driver, low number of printable materials for AM/3D printing. Some of the
sub-drivers were suggested in the context of more than one driver; data security risk
e.g. was suggested in the context of Big Data/AI, Cloud and IoT. Sub-drivers with the
same content could have been suggested both as inhibitors or accelerators, but that
did not happen.
Some suggestions of sub-drivers formed almost Hegelian thesis-antithesis relation-
ships, like the one relating whether users open their data (i.e., open = accelerates; not
open = inhibits). The same example can also be characterized as following Boolean
logic even in individual use cases; the user opens its particular section of the machine
data or not; the user opens all of its machine data or not. Following the user data ex-
ample, the likely expert view for that sub-argument in the Delphi round two is a heuris-
tic judgment of the average state of affairs; majority (minority, most, only some …) of
the users are likely to refrain from giving access to their data based on my judgment of
the benefits of the technological driver that I am aware of. It appeared that the inhibitive
sub-drivers were more often attached to behavioral aspects (like beliefs), whereas ac-
celerative sub-drivers were easier to quantify (e.g., short development time).
More than 30 inhibitive and 50 accelerative sub-drivers were proposed. All those that
were named more than once were merged onto the list from literature and research
reports. In that merger, sub-drivers that had the same content but different names were
placed under one name. The merged list of technological drivers and sub-drivers is
presented in Table 6.
111
Inertia sub-driver Accelerator sub-driver
AM / 3D
printing
Limited size of the object
Slow printing speed
Low # of printable materials
High investment cost
Lack of competences
Management beliefs
Investment in current infrastructure
Digital distribution (e.g. spares)
Manufacturing recipe as IPR
Unique mass production
Sustainability
Manufacturing close to usage
New product features
Big Data / AI Lack of relevant history data
Data security risks
Data ownership confusion
Data ownership conflict of interest
Management beliefs
Lack of competences
Privacy concerns
Reduce of human errors
Knowledge scalability
Diagnose speed
Prescriptive potential
Core element in digital B2B business models
Emergence of platform tools
Advancement in cognitive computing
Users open their data
Cloud com-
puting
Data security risks
Investment in legacy systems
Lack of competences
Enabler for new business models
Quantum leap in operational productivity
Emergence of platform solutions
Low cost of computing
High supply of partners
Lower barrier of computing
Internet of
Things
Data security risks
Lack of standardization
Lack of systemic understanding
Lack of competencies
Non sensored Installed fleet
Cost of data transmission
Availability of data transmission
Management beliefs
Systemic productivity
New source of customer insight
Emerging platforms
Digital entrants to equipment business
Inceptive to crowd
Enabler of new business models
MBSE Incompatible legacy systems
Legacy processes
Lack of competencies
Management beliefs
Reduced # of errors in development
Short development time
Low development cost
Enabler of crowd knowledge
Enabler of customer engagement in develop-
ment
Robotization Legislation lagging behind
High capital cost
Lack of infrastructure
Social disruption
Technological risks
High switching cost
Manned design paradigm
Systemic productivity
Low wear, tear and collision costs
Good LCC predictability
High safety
Sustainability
Compensation of labor shortage (incl. labor
unrest)
Table 6. Results from Delphi round one, technological drivers
112
The process of validating the economic / strategic management drivers was the same
as with the technological drivers. As described in the methodology chapter, the valida-
tion included only those two drivers that are directly related to digitalization (Digital
Business Models and Networks, Crowds and Platforms). The other five drivers of the
Figure 11 were not included in the validation because they were either seminal theories
used to explain industry evolution or fundamental components in economics, such as
firms and capital.
Out of the 24 interviews, 13 views were given that the either of the two proposed driver
could be disruptive in the context of the research question, 14 that the maximum that
the driver could reach is a quantum leap in competiveness, and only three that viewed
the driver as unable to achieve more than continuous improvement. Based on the re-
sults both drivers were included to the list of other five economic / strategic manage-
ment drivers as illustrated in the Figure 11.
As a note, the round one interviews also yielded comments about Convergence, Radi-
cal Innovation, Five Forces, Stakeholders, and Capital. Convergence was mentioned
several times to explain strategic management mechanisms and to explain potential
disruption, which endorsed the argument for it to be in the list for Delphi round two.
Radical Innovation as a disruptor concept was viewed to be based on the content of
the innovation itself. Capital discussion was pre-dominantly around the potential impact
of “Silicon Valley” types of Venture Capital on the asset-heavy industries like machine-
building. The role of capital more often was viewed as a catalyst of a disruption that
has already started, rather than the initiator of that process.
Just as with the technological drivers, the majority of the interview time was spent on
discussing sub-drivers that can either inhibit or accelerate the potential disruptive im-
pact of the economic/strategic management drivers. The proposed sub-drivers were
merged with the sub-drivers of the literature, the same way as with technological sub-
drivers, and the results appear in Table 7.
Inertia sub-driver Accelerator sub-driver
Five Forces
(six)
Rivalry between existing competitors
Cost driven customers
High entry barrier industry
Long established stakeholder organiza-
tions
Oligopolistic market
Digital entrants to traditional industries
Substitutive business models
Value driven customers
Value driven suppliers
Influential Early Adopters
Suppliers passing middle-men
Active 6th force
113
Inertia sub-driver
Accelerator sub-driver
Industry con-
vergence
Industry organs
Industry beliefs, strong dominant logic
Cultural barriers
Industry specific financial KPIs
Platform supply
Management rotation
Business model diffusion
General digital technologies
Consumer business behavior into B2B
Radical inno-
vation
Legacy competences
Legacy processes and systems
Legacy leadership and management
Return on continuous improvement
Risk adverse business culture
Conservative industry
Lack of absorptive capacity
Software intensity
Access to crowd knowledge
Experiment culture
Influential Early Adopters
New level of customer insight
Powerful development tools
Digital busi-
ness models
Data security risks
Data ownership confusion
Fixed legacy systems and processes
Lack of dominant theory
Incumbents active resistance
Lack of trust in virtual environments
High barrier industry
Quantum leap in productivity
Real time opportunity
Asset light
Influential Early Adopters
Slow growth of existing business
Startup emergence
Networks,
Crowds,
Platforms
Lack of capabilities
Management beliefs
Lack of systems and processes
Lack of trust in virtual environments
Conflict in IPR
Platform availability (discipline)
Unrestricted capacity
Fixed cost light
Startup emergence
Diginatives to firms
Open source hardware
Self-fulfillment motives of knowledge workers
Crowd intelligence effect
Firm, Strate-
gy, Manage-
ment
Strong Industry beliefs
Next quarter pressure
Leadership from the past
Well established dominant logic
Complex organizations
Hierarchical organizations
Incrementalism
Bandwagon, Hype, Halo
Management consulting focus
Slow growth of traditional business
Management rotation
Emerging digital strategy frameworks
Experiment culture
Degree of external engagement
Risk capital Missing financial track record
Inexperienced management
Impatience
Crowdfunding
Asset light
High upside
Low interest rates
Success stories
Table 7. Results from Delphi round one, economic / strategic management drivers
114
Additional observations from the Delphi round one
The World Economic Forum (2015) suggested a path to Outcome Economy that was
seen in Figure 5. That view was also similar to what Porter and Heppelman (2014) de-
scribed—from products to systems of systems. The first observation relates to those
concepts from several perspectives in the free comments during the Delphi round one.
There is a high complexity of optimizing components (diesel engine) or ma-
chine performance because it consists of so many subsystems;
A small amount of critical knowledge from upper levels has high value leverage
if it can be coupled to lower lever development;
Decoupling of level or decoupling of value chain is always a potential source of
disruption;
Decoupling opportunity can attract out-of-industry players (DHL starting to print
some of the product as a service was named as example of decoupling part of
the physical supply chain and while still satisfying the need of physical goods);
Creation of a higher level delivery has an advantage of little to no competition
(Kim and Mauborgne, 2005), but also has a risk if customers of the current val-
ue chain are not optimal or ready for the new type of offering.
Those types of comments that arose from the first round Delphi interviews were used
to construct a preliminary framework that looks at disruption the same way as the
World Economic Forum report (2015) or Porter and Heppelman (2014), but suggests
the paradox of complexity and value if the layers can be decoupled. The preliminary
framework is illustrated in Figure 14, which also provides guidance to the case design.
Figure 14. Preliminary finding on systemic characteristics of disruption
Component
Machine
Fleet
Process
System
New actors
115
The second observation was the emergence of four groups of concepts. Cloud, MBSE,
and AM/3D printing were perceived as tools in digitalization. They were powerful, with
different features attached to different applications, but ultimately they were perceived
(according to some of the comments) as commodities available for everyone. Iot and
Robotization were perceived as the second group -phenomena: connectivity and au-
tonomy, respectively. Therefore, it was suggested that they might contain true disrup-
tive nature. The disruptive nature of Networks, Crowds, Platforms, and Big Data/AI, on
the other hand, differed. Crowds and platforms were suggested to be one feature of a
network economy paradigm, moving the power to crowds and users. This shift of power
was suggested to create a leadership challenge for the players in the current value
chains. The most important discussions about Platforms was whether they will be verti-
cal or horizontal, and how many Platforms there can be in without losing the benefits of
the network effects.
The third group, Big Data/AI, in this context appeared to be its own category as the key
enabler. Many comments, like “the business is getting data-centric,” were mentioned.
Big Data was only partly attached to AI and machine learning. AI (together with AM)
was expected to still have strong technological advancement, and the roles of algo-
rithms and software as the carriers of value were mentioned. The fourth group was the
value model or business model. Industrial Internet was mostly referred to as a con-
cept without generic meaning, whereas Industrial Internet as defined by General Elec-
tric was referred to as one Digital Business Model. Another specific Digital Business
Model suggested was an IPR fee for printing spare parts.
The third observation was related to comments about timing. The most distinctive it
was with tools and the key enablers. Cloud and Big Data were deemed to be technol-
ogies that are already matured. MBSE was referenced as a methodology/tool that still
is advancing, but whose current usage is below the current potential. AM/3D printing
and AI were expected to have the largest disruptive potential depending on technologi-
cal advancement. There were also some comments on the objectives of the impact.
Operations and products were more often mentioned in a quantum leap context, and
business models in disruption. Disruption in service meant both operational discontinui-
ties and the changes of business models.
116
4.2.2 Second Delhi interview round
The number of panelists increased from 24 to 36 from the first Delphi round. They rep-
resented the 3x3 background matrix evenly, which meant that each role/value chain
cell contained four panelists. Five panelists from round 1 were also interviewed at the
second round. The second round was represented by nine different universities or re-
search institutes and 22 companies, of which 19 were in global business. Two compa-
nies and three universities had more than one panelist, but not from the same cell in
the matrix. Panelists came from 11 different countries, and 60% were not Finns.
The first objective of the Delphi interview round two was to rank order technological
and economic/strategic management drivers from the Delphi round one. The second
objective was to find possible new inhibiting and accelerating sub-drivers, and then
rank order the three most important ones. The third objective was to detect differences
on views on the drivers and sub-drivers between enablers, actors and users. As a last
analysis there is a grouping of the same kind of sub-drivers that are either inhibiting or
accelerating technological or economic / strategic management drivers as whole.
4.2.2.1 Technological drivers
The results of the order ranking of the technological drivers appear in the table 8.
Driver Number 1
rankings
Average
ranking
Variance
Big Data / AI 18 1,9 1,3
IoT 9 2,4 1,6
MBSE 5 3,9 3,0
Robotization 2 3,7 2,3
AM / 3D printing 2 4,2 1,7
Cloud computing 0 4,9 1,7
Table 8. Rank order of the technological drivers in Delphi round two
117
The results in the Table 8 represent the rankings of the panel as whole. Figure 15 pre-
sents the results visually so that each cell consists of four panelists’ three highest
ranked technological drivers and different colors represent technological drivers. As an
example, one enabler/ decision-maker ranked Big Data / AI as number 1, two enabler/
decision-makers ranked IoT as number 1 and one enabler/ decision-maker ranked
MBSE as number 1 as potential technological disruption driver. The same logic applies
to number 2 and 3 rankings and for all nine cells.
Figure 15. Three highest ranked technological drivers in the Delphi panel in round two
In the third analysis, the technological drivers are aggregated by the value chain, as
shown in Figure 16. Each driver shows how many number 1, 2 and 3 rankings
1 2 3
Actor / Decision maker
1 2 3
Actor / Scientist
1 2 3
Actor / Syntheziser
1 2 3
User / Scientist
1 2 3
Enabler / Decision maker
1 2 3
User / Syntheziser
1 2 3
Enabler / Scientist
1 2 3
Enabler / Synheziser
AM / 3D printing
Big Data / AI
Cloud computing
Internet of Things
MBSE
Robotization
1 2 3
User / Decision maker
Note: Y- axes is number of respondents and X – axes is rating.
118
(marked with colors) it got from respective value chain members, like the first column
in the figure indicates that AM/3D printing got one number 2 ranking from the enablers
and two number 3 rankings from the enablers. The Figure 16 indicates that Big data/AI
and IoT received support independent from their places in the value chain. MBSE had
commonalities with enablers and actors, and Robotization had the same kind of pattern
between actors and users.
Figure 16. Three highest ranked technological drivers in the Delphi panel in round two by the value chain
When the participants were assessing the three most critical technological drivers, they
came up with 30 new sub-drivers that either inhibited or accelerated the driver. Table 8
illustrates all the new sub-drivers suggested, listed in the order of appearance (i.e.,
interview number), including the main driver, inertia (I) or acceleration (A), and how
many order rankings (1, 2, or 3) the sub-driver achieved after the appearance. The
growth of sub-drivers started to saturate in the middle of the interview round, and
turned more sporadic thereafter. The order ranking of new sub-drivers is analyzed
based on its pattern in order to judge whether the result would have been different if
the sub-drivers would have been available from the beginning.
0
2
4
6
8
10
12
3
2
1
Note: Y- axes is number of respondents and X – axes is the place in the value chain/ driver
Colors indicate rating
xes is rating.
119
Inter-
view
Sub-driver Driver I A 1 2 3
1
1
2
5
5
6
7
8
9
11
11
11
12
13
13
13
13
14
16
16
16
17
18
19
20
21
21
22
29
30
High product cost
Platform tools
Logistics companies start printing
Ethical considerations
Market for intelligent data missing
Advancement in 3D technology
Legislation lagging behind
Lack of systemic understanding
Systemic productivity
Industry beliefs
Lack of business model change
Drives collaboration
Legacy knowledge in systems
Lack of use cases
Cost of technology
Hype (unrealized)
Safety
Market for intelligent data emerging
Deployment risk
Lack of standardization
Delivery capabilities
Lower cost of analytics
Lack of competence
Power of visualization
Advancement in user experience
Silo organizations
Data ownership conflict of interest
Real time knowledge
Used cases
Technology becoming mature
AM / 3D
MBSE
AM / 3D
Big Data/ AI
Big Data/ AI
AM / 3D
IoT
MBSE
MBSE
Big Data/ AI
Cloud
Cloud
Cloud
IoT
IoT
Big Data/ AI
Big Data/ AI
Big Data/ AI
Robotization
Robotization
Robotization
Big Data/ AI
Robotization
MBSE
IoT
MBSE
IoT
IoT
Big Data/ AI
IoT
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
1
1
1
3
2
2
1
3
1
1
3
1
1
1
1
2
3
2
3
1
4
1
2
1
1
1
1
1
1
1
1
3
1
1
1
5
1
1
4
3
1
1
3
1
1
2
1
2
Table 9. New technological sub-drivers from the Delphi round two
Sub-driver analysis is done for the top three drivers from Table 8 by presenting first the
three most important inhibitive and accelerative sub-drivers based on the view of the
panel as whole. The ranking of sub-drivers was done by giving three points to each
sub-driver that was ranked as number 1, two points for number 2 ranking, and one
120
point for number 3 ranking. The analysis states the percentage from all points the sub-
driver received. Some of the closely related sub-drivers are clustered.
After ranking sub-drivers are elaborated with respective free comments that arise in the
Delphi interview round two. As a final sub-driver analysis the degree of consensus in
the value chain is evaluated. The sub-drivers of the three lowest ranked drivers are
analyzed as group at the end.
Big Data/AI
Three clusters of inhibitive sub-drivers got 79% of the points for Big Data/AI:
Beliefs, 26% (management and industry beliefs)
Data 31% (confusion and conflicts of ownership, security)
Competence, 22%
Industry beliefs, which came from the interview number 11 was the only new sub-
driver that could have had any material difference in analysis. It collected eight percent
of the points in the consecutive interviews, and beliefs as an inhibitive driver probably
would have been even stronger if industry beliefs had been available from the begin-
ning.
The accelerative sub-drivers for the Big Data/ AI were not clustered, and five individual
sub-drivers achieved 75% of all points:
Prescriptive potential, 22%
Knowledge scalability, 15%
Users opening their data, 14%
Advancement in cognitive computing, 12%
Reduction of human errors, 12%
Twenty-seven out of 36 participants elaborated on the importance of Big Data/AI in the
disruption during their free comments on the choice of order of the sub-drivers from
Table 6 or 9. Three new generic topics emerged from those comments: complexity,
progress, and content. Increasing complexity was mentioned in several interviews, and
it was described as the need for simultaneous optimization in several levels of a sys-
tem that are also interlinked. This development was perceived to exceed the human
ability to cope with it, and therefore tools like Big Data were described to have disrup-
tive potential. Big Data was mentioned as one contributor to AI development, but at the
121
same time it was acknowledged that AI is still in its infancy stage compared to its po-
tential to disrupt industries. The key individual point in AI was that exception handling is
still dominated by human judgment. Free comments related to development of machine
learning and adaptive systems came more from the AI driver discussion than from ro-
botics. The third emerging Big Data/AI topic, content, stemmed from an industry or dis-
cipline perspective. “General industry Google” was not perceived likely, but instead
many comments about discipline-based, data-driven platforms were made. This was
coupled with the scalability (e.g., the friction of bearings in rotating machines as a
source for a data-driven niche platform). The same justification was made for data-
driven industry platforms. One comment summarized the content view: “All the right
answers are today accessible to anyone; you have to know the right question, which
ability comes from the content knowledge.”
Consensus is analyzed so, that if any of the value chain member’s percentage of points
of the sub-driver differs more than five percent units from the average percentage of
that sub-driver, it implies non-consensus.
In the inhibitive drivers, there was non-consensus in the following sub-drivers (clusters):
Data (E= 35 %, A= 22 %, U=39%)
Competence (E= 23%, A=26%, U=17%)
The non-consensus in accelerative sub-drivers was with:
Prescriptive potential (E=18% %, A=17 %, U=29%)
Knowledge scalability (E=11%, A=24%, U=11 %)
Advancement in cognitive computing (E=24%, A=6%, U=5%)
Internet of Things
Four inhibitive sub-drivers or clusters got 81% of the points:
Lack of systemic understanding/ competence, 38%
Management beliefs, 17%
Lack of standardization, 15%
Data security, 11%
122
The lack of used cases got seven percent of the points even though it appeared in in-
terview number 13; those points were mainly away from the lack of standardization.
Three accelerative sub-drivers got 75% of all points:
Enabler of new business models, 27%
Systemic productivity, 23%
New source of customer insight, 19%
The cost of technology got eight percent of the points, and because it appeared in in-
terview 13, it would have attracted more points, which would not have been taken away
from any specific alternative. There was no any material new information about IoT in
the free comments.
In the inhibitive drivers, there was non-consensus in the following sub-drivers:
Lack of systemic understanding/ competence (E= 29%, A=28%, U=57%)
Lack of standardization (E=17%, A=23%, U=4%)
The only non-consensus accelerative sub-driver was the new source of customer in-
sight (E=30%, A=13%, U=11%).
Model Based System Engineering
MBSE got its position in the final top three due to receiving more number one rankings
than Robotization. The other justification for its high ranking came from free comments,
where a number of participants stated that the disruptive part of Robotization is in AI,
machine learning, or algorithms, rather than machines getting smart enough to operate
without drivers.
Three inhibitive sub-drivers or clusters got 88% of all points in MBSE:
Lack of systemic understanding/ competence, 49 %
Legacy processes, 24%
Management beliefs, 15%
Three accelerative sub-drivers or clusters got 77% of all points in MBSE:
123
Productivity (systemic or based on the shorter lead times), 37 %
External engagement (crowds and users), 27%
Power of visualization, 13%
Visualization became available in interview 27 but still rated 13%, which likely have
been higher if it had been included from the beginning.
The free comments yielded some additional viewpoints to MBSE. Most notable was
that modeling is one tool to manage the increasing complexity in AI, which benefits the
current tools as well as accelerates the utilization of AI. It was also referred to as an
enabler of simulations and higher level algorithms. The second topic was about en-
gagement, both internally and externally, and MBSE was mentioned as a platform that
uses visualization and simulations as methods. MBSE was the only technological factor
that had role-based patterns where the benefit was more seen by synthesizers (and
somewhat by scientists), regardless if they were enablers, actors, or users. A few of the
interviewees did mention that design systems are lower-cost investments than ERP
systems, and therefore most of the decision-makers have little understanding of the
benefits. One interviewee also stated that modeling the current knowledge is the best
investment both in the short and long term.
In the inhibitive sub-drivers, there was non-consensus in the following sub-drivers of
MBSE:
Legacy processes (E=8%, A=29%, U=33%)
Management belief (E=33%, A=10 %, U=6%)
The non-consensus in accelerative sub-drivers was with:
Productivity (E=46%, A=40%, U=17%)
External engagement (E=21%, A=24%, U=44%)
Power of visualization (E=21%, A=7%, U=17%)
AM/3D printing, Cloud computing, and Robotization
AM/3D printing got two number one ratings as seen in the Table 8. Almost half of the
panelists acknowledged in their free comments that AM might be, for industrial produc-
tion, the most prominent disruptive technology when the technological barriers have
124
been solved; however, during the interview, respondents did not rank it higher as the
timeline was uncertain and longer in the future. The inertia in the sub-driver ranking
was predominantly divided into two categories: technical immaturity for large scale pro-
duction (size, material, speed), and management beliefs. One panelist expressed a
view that there will be hybrid technologies (conventional attached to AM) during the
evolution. The common denominator for the other free comments could be described
as weak signals. Comments served also as concrete explanation to elaborate the
choice of more abstract arguments (drivers or sub-drivers) as stated in methodology
section 3.3.1. These types of comments were:
“DHL has started to use 3D printing as a tool in their delivery.”
“In spring 2015, 50% of recruitment in manufacturing in US required 3D skills.”
“3D revolution is not driven by a paradigm of how you print existing designs but
how you design when things can be printed.”
“3D printing is a powerful technology to impact current trade imbalances.”
Cloud had the highest consensus among the technological factors. It was perceived as
the dominant technology in computing, but as a standalone was not disruptive. Even
those few who lifted it into ranks 1-3 viewed the disruption to only be based on simulta-
neous business model innovation. Also, the ones who did not rate Cloud highly
acknowledged that it has changed the paradigm of computing (e.g., cost, availability to
anyone, scalability). Some saw security as inertia; others viewed that world class Cloud
suppliers provide better security than even good companies can achieve internally.
Robotization or autonomous processes had medium disruptive results. The dominant
view of robots in the industrial context was evolutionary, and Robotization was seen as
technical progress for productivity. This was supported by the sub-driver analysis. The
inertia was technical, (23%, technology or deployment), financial (29%, capital or
switching cost), or social disruption (18%). Social disruption in the technological devel-
opment has been an issue for as long as technology-based productivity has replaced
human work for different jobs; this disruption is both a leadership and cost issue. Also,
the accelerative sub-drivers of Robotization were similar to past automation develop-
ment (productivity 32%; addressing labor cost of availability 25%). Ten panelists ex-
pressed free comments on Robotization; five of the comments were pondering the
question whether the end game is intelligent robots working together or slaves to a
smart Cloud. There was not a strong view on this idea, but it was perceived as a vital
question for development. Three panelists emphasized the medium steps in automa-
tion, like remotely operated machines, where the machine is intelligent but the planning
125
and exception handling is managed by a human operator. One panelist, however,
pointed out that mixing manned and unmanned operations creates new types of securi-
ty risks. The remaining comments were related to changing industrial job profiles, from
field to office.
Interrelated technological sub-drivers
The last analysis was trying to detect sub-drivers that impact to several technologies.
This is achieved by three steps: first counting points of the exact same sub-drivers
(such as data security), then analyzing whether there are easily available a more ge-
neric grouping where more than one sub-driver can be combined; low # of printable
material in AM and technological risks in Robotization can be categorized technical
barrier as an inhibitive sub-driver. The final step was counting the points based on the
new grouping. The new aggregated technology related sub-drivers can be seen in the
Figure 17 in rank order; eight aggregated sub-drivers constituted 92% of the inhibitive
sub-drivers and 91 % of the accelerative sub-drivers. The summary of the results of
technological drivers is shown in Figure 18.
Figure 17. Top eight aggregate technological sub-drivers
0 20 40 60 80 100 120 140 160 180
Lack of competences / understanding
Productivity leap
Data (security, quality, availability, ownership)
New insight
Management / Industry beliefs
New business models
Technical barriers
Platforms and networks
Legacy (systems, processes, infrastructure)
Cost
Legistlation, standardization
Advancement in technology
Cost (investment, product)
Quality
Social , Ethics
Speed
Number of points based on the ranking Red = Inhibiting, Blue = Accelerating
126
Big Data/AI Management/Industry beliefs Prescriptive potential
E A U Data (security, confusion, owner-
ship)
Knowledge scalability
Competence Users opening their data
IoT Lack of systemic understanding
and competence
Enabler of new business models
E A U Management beliefs Systemic productivity
Standardization New source of customer insight
MBSE Lack of systemic understanding
and competence
Productivity (including lead
times)
E A U Legacy processes Enabler of external knowledge
Management beliefs Power of visualization
Figure 18. Summary of the technology drivers by the value chain
Note: The colors represent consensus. Green means consensus with the value chain mem-
bers who have green color. If one of the three disagrees, red means that the one who disa-
grees has stronger support from the green ones and weaker support from the yellow ones. In
the driver section, E = Enabler, A = Actor, and U=User. With sub-drivers, the logic is that the
bottom square is Enabler, the middle square is Actor, and the upper square is User. The
inhibitive sub-drivers are located in the left column and the accelerative sub-drivers in the
right column.
127
4.2.2.2 Economic / strategic management drivers
The results related to economic / strategic management drivers have been analyzed
with the same procedure as the results related to technological drivers. They are also
presented the same way as the technological drivers in the previous sub-chapter.
Therefore, the procedures of the analyses, or instructions how to read the results are
not repeated in this sub-chapter. The results of the order ranking of the economic /
strategic management drivers appear in the table 10 representing the rankings of the
panel as whole.
Driver Number
1 rank-
ings
Average
ranking
Variance
Networks, Crowds, Platforms 13 2,2 1,4
Convergence 9 3,8 4,4
Radical Innovation 7 3,8 3,3
Digital Business Models 4 2,7 1,4
Firm, Strategy, Management 2 4,1 1,7
Six Forces 1 5,2 2,7
Risk Capital 0 6,4 1,5
Table 10. Rank order of the economic / strategic management drivers in Delphi round two
Figure 19 presents results visually so the order ranking of each of the nine cells can be
seen separately.
128
Figure 19. Three highest ranked economic / strategic management drivers in the Delphi panel in round two
The aggregation of the economic/strategic management drivers by the value chain can
be seen in Figure 20. There was a high consensus on the importance of Network,
Crowds, and Platforms. Similar consensus existed with Digital Business Models, in-
cluding agreement that its disruptive nature was less than that of Networks, Crowds,
Platforms. Convergence and Radical Innovation formed the only material differences
by the value chain. Both were perceived important, but Convergence was more favored
by enablers and actors, whereas Radical Innovation was favored by enablers and us-
ers.
Six Forces
Industry Convergence
Radical Innovation
Digital Business Models
Firm, Strategy and Management
Risk Capital
Networks, Crowds, Platforms
1 2 3
Enabler / Decision maker
1 2 3
Enabler / Syntheziser
1 2 3
Enabler / Scientist
1 2 3
Actor / Decision maker
1 2 3
Actor / Syntheziser
1 2 3
Actor / Scientist
1 2 3
User / Decision maker
1 2 3
User / Syntheziser
1 2 3
User / Scientist
Note: Y- axes is number of respondents and X – axes is rating.
129
Figure 20. Three highest ranked economic / strategic management drivers in the Delphi panel in round two by the value chain
The panelists suggested 31 new economic / strategic management sub-drivers that
can be seen in Table 11.
Inter-view
Sub-driver Driver I A 1 2 3
1
1
5
7
7
7
11
11
13
14
Long term orientation
Diginatives to firms
Privacy concerns
Lack of domain knowledge
Different product cycles
Platform availability (industry)
Successful history
Visionary leadership
Ability to move management be-tween industries
Legacy systems and processes
Radical Innovation
Firm, Strategy, Management
Networks, Crowds , Platforms
Convergence
Convergence
Networks, Crowds, Platforms
Firm, Strategy, Management
Firm, Strategy, Management
Risk Capital
Networks, Crowds, Platforms
x
x
x
x
x
x
x
x
x
x
1
2
1
2
5
2
2
2
1
1
5
2
3
2
4
5
0
2
4
6
8
10
12
3
2
1
Note: Y- axes is number of respondents and X – axes is the place in the value chain/ driver
Colors indicate rating
xes is rating.
130
Inter-view
Sub-driver
Driver
I
A
1
2
3
15
15
17
19
19
21
21
21
21
26
26
27
28
31
Incumbents active resistance
Software intensity
Lack of intelligence in algorithms
New level of customer insight
Client vs. market focus
Silo organizations
Pressure from financial market
Pressure from financial market
Use of several platforms
Based on small invoices
Disappearance of distance
“Conservative users”
Lower transaction cost
User engagement
Networks, Crowds, Platforms
Digital Business Models
Digital Business Models
Digital Business Models
Firm, Strategy, Management
Digital Business Models
Digital Business Models
Firm, Strategy, Management
Networks, Crowds, Platforms
Digital Business Models
Firm, Strategy, Management
Digital Business Models
Networks, Crowds, Platforms
Networks, Crowds, Platforms
x
x
x
x
x
x
x
x
x
x
x
x
x
2
3
2
1
1
1
1
1
2
2
1
3
1
2
2
2
1
1
1
Table 11. New economic / strategic management sub-drivers from the Delphi round two
The rank order of the Table 10 for the sub-driver analysis of economic / strategic man-
agement drivers is altered to be for the top three: 1. Networks, Crowds, Platforms; 2.
Digital Business Models; 3. Industry Convergence. The justification for Digital Business
Models moving upward is the second lowest average ranking and equally low variance
as with the number one, Networks, Crowds, Platforms.
Networks, Crowds, Platforms
Three inhibitive sub-drivers constituted 72% of points for Networks, Crowds, Platforms:
Management beliefs, 26%
Lack of capabilities, 23%
Systems and processes (both legacy and lack of new), 23%.
The late sub-drivers’ legacy systems and processes (came from the interview number
14 and collected 11 % of the points) and incumbents’ active resistance (came from the
interview number 15 and collected five percent of the points) were the only ones who
could have impacted to the analysis. The points of sub-driver, legacy systems and
processes were later mainly away from lack of processes, which is still a similar sub-
131
driver and thus was aggregated. Incumbents’ active resistance sub-driver started to
reduce points from both of the process sub-drivers.
Three accelerative sub-driver clusters for the Networks, Crowds, Platforms achieved
77% of all points:
Emergence of platforms (discipline and industry), 32 %
Cost (fixed cost, transaction cost, capacity), 24%
Crowd intelligence effect (including users’), 21%
Thirty-one out of 36 participants ranked Networks, Crowds, Platforms into top three and
26 participants also elaborated on the topic further in their free comments. Other than
the original or emerged sub-drivers, free comments fell under four different themes:
diversity, users, eco-systems, and culture. Diversity was strongly attached to the exter-
nal knowledge in innovation. The comments highlighted that even large companies
have many subject matter resources; group thinking might lead to low diversity, which
can be mitigated by the methods of Open Innovation. Users’ opening their data was
described as a disruptive enabler when users open their data beyond suppliers. Eco-
systems were discussed in two contexts: they are very different from the value chain,
and managing them requires very different management styles and capabilities. Cul-
ture was also referenced from two very different angles. Some comments were close to
management beliefs and related to the internal culture as a barrier to benefiting from
the external world (other than directly from the value chain). The other cultural com-
ments were more philosophical, posing the question of whether openness and trans-
parency as concepts are shared in different national cultures, which most likely they
are not—a relevant issue for multinationals. The free comments also contained a num-
ber of direct views on the open/close paradigm, and they were similar in tone: “Open
systems will, at the end, win.”
In the inhibitive drivers, there was non-consensus only in lack of capabilities (E= 27 %,
A= 27 %, U=13%).
The non-consensus in accelerative sub-drivers was with:
Emergence of platforms (E=28%, A=24%, U=45%)
Crowd intelligence effect (E=8%, A=29%, U=18%)
132
Digital Business Models
Three inhibitive sub-drives or clusters got 78% of all points in Digital Business Models:
Industry barriers (including incumbents’ active resistance), 37%
Legacy systems and processes, 23%
Data (security and ownership), 18%
Silo organizations first came up in interview 21 and started to collect points thereafter,
ending up with 9% of the points. Silo organization points were away from industry bar-
riers. This suggests that silo organizations as a sub-driver might have been in im-
portance close to those in the top three if present from the beginning.
Three sub-drivers or clusters constituted 72% of the accelerative sub-drivers:
Cost benefits (productivity, capital), 33%
Influential Early Adopters, 17%
Slow growth of existing business, 15 %
New levels of customer insight, 12%
New levels of customer insight originated in interview 19 and collected 12% of the
points, mainly from productivity; this suggests that it would have ended higher in the list.
From the free comments, new perspectives were found to form 4 themes: winners,
timing, supply, and used cases. The winners discussion pondered the issue if the
startups could win incumbents by using Digital Business Models. The challenge was
perceived to be in high industry barriers and in the need for assets. Progressive incum-
bents or adjacent systems suppliers (e.g., type GE; ABB, Siemens) who possess some
of the critical technologies were named also as potential disruptors. The need for both
digital and domain capability (as with AI) was named as a prerequisite to winning with
Digital Business Models.
Timing comments mostly supported the opinion that there will be a pivot point for the
growth of Digital Business Models; forecasting the exact timing, however, is not likely.
The essence of the supply comments was that Digital Business Model in machine-
building is currently more marketing than supply. Therefore, the importance of the used
cases is high, and was elaborated upon in many discussions. This is a bit of a Catch-
22, as everybody hopes for the occurrence, but nobody is willing to move. One specific
133
view was that an oligopolistic market structure may lead users to attract new suppliers
with Digital Business Models in order to enhance the competition.
In the inhibitive sub-drivers, there was non-consensus only in data (E= 13 %, A= 17 %,
U=24%).
In the accelerative sub-drivers, there was non-consensus in:
Cost benefits (E=20 %, A=43 %, U=35%)
Slow growth of existing business (E=19 %, A=22 %, U=6%)
Industry Convergence
Four inhibitive sub-drives got 83% of all points in Industry Convergence:
Industry beliefs (includes Dominant Logic), 32%
Industry-specific financial steering, 22%
Cultural barriers, 15%
Lack of domain knowledge, 14%
Three accelerative sub-drivers got 88% of all points:
Business model diffusion, 41%
Consumer business behavior merging into B2B industries, 32%
Generic digital technologies,15%
From the free comments, some themes were repetitive: speed, platforms, consumers,
and individual companies. Speed was related to industry beliefs and specifically to
clock speed; most examples were phrased like: “What is fast development time – real
time or 2 years?” Platforms were attached to generic digital technologies from a disrup-
tion perspective—especially discipline-based. Also, industry overlapping ecosystems
were mentioned. Most of the consumer elaboration related to application business con-
cepts (apps) or rising user interface expectations in B2B driven by the people’s private
experiences. Individual company initiatives, like earlier examples of DHL in 3D printing
or Amazon in logistics, were referenced in industry convergence.
134
In the inhibitive sub-drivers, there was non-consensus in the following sub-drivers of
Industry Convergence:
Industry beliefs (E=36%, A=35%, U=17%)
Industry-specific financial steering (E=19%, A=27%, U=11%)
Cultural barriers (E=14%, A=13%, U=22%)
Lack of domain knowledge (E=17%, A=6%, U=28%)
In the accelerative sub-drivers, there was non-consensus only in business model diffu-
sion (E= 36 %, A= 48 %, U=33%).
Radical Innovations; Firm, Strategy, Leadership; Industry forces; Risk Capital
Free comments indicated that disruption by Radical Innovations is more isolated events
than a result of a predictable pattern. Within these limitations, the inertia in Radical In-
novation was mostly explained by 3 sub-drivers: legacy leadership and ownership
(26%), good returns by continuous improvement (21%), and risk-adverse business
culture (19%). The sub-drivers that accelerate Radical Innovation were experiment
culture (29%), influential Early Adopters (21%), and the new level of customer insight
(19%). The dominant view in the free comments resembled Christensen’s Innovators
Dilemma (2011): continuous improvement is a low risk activity which almost always
pays back. Another view was that digitalization increases raw material for Radical In-
novation (e.g., sensors, computing, behavioral data, opinion data), which theoretically
increases the chance of Radical Innovation as an event. There were also reminders
that Radical Innovations are often low tech. The last category of comments related to
business definitions or business models, and those comments emphasized that part of
radicalness is redefinition (e.g., moving into higher systemic levels, as discussed earlier;
or Blue Ocean type (Kim and Mauborgne, 2005) of thinking as discussed in the theory
section).
Free comments on Firm, Strategy, Management were similar to those of Radical inno-
vations, and centered on the inhibiting side around Christensen’s (2011) view, and on
the acceleration side around “Steve Jobs” (i.e., was the smart phone disruption be-
cause of Steve Jobs, the product, Apple as a company, or its strategy and capability to
implement it). However, the overriding perspective was the same as in Radical Innova-
tion: the firm is at the core of the changes inside of an industry, but is rarely a disruptor
135
of the industry as a whole. One comment sums up the average: “Large companies do
disruptive things only when forced to do so.” Sub-drivers that slowed down Firm, Strat-
egy, Management as a disruptive driver were: successful incrementalism (28%), firm
itself (e.g., authorial leadership and hierarchical, complex organizations; 23%), industry
beliefs (23%), and short-term financial pressure (22%). The accelerative sub-drivers
were somewhat the opposite: Visionary leadership (33%), experiment culture (23%),
and slow growth of existing business (18%).
Industry Forces was an interesting driver; practically every panelist recognized it as
“The Theory” explaining industry dynamics. The recognition was similar to stakeholder
theory, and still the combination ranked second lowest as an economic/ strategic man-
agement driver to explain disruption by digital technologies. The dominant inertia sub-
drivers were high-entry-barrier industry (35%) and cost-driven customers (31%),
whereas three top accelerative sub-drivers were: Influential Early Adopters (37%), val-
ue driven customers (24%) and substitute business models (20%). One group of free
comments was pondering the same question as in Networks, Crowds, Platforms: that
the value chain is not equal to ecosystem, and the Industry Forces are strongly linked
to the value chain. This was perceived as one weakness of the theory to explain net-
work-based disruptions. The inertia of industry barriers was explained with the view
that the substitute suppliers in consumer business are often startups, but in high-barrier
industries they are not credible. The conflicting views propose that niche startups can
steal pieces of value from incumbents, or—when united—even bigger pieces. Some
comments stated that the highest value generation potential is not constant to any
place in the value chain. Theoretical views suggested that Industry Forces explain S-
curve leaps inside industries but not transformation between industries or systemic S-
curve leaps. There were also comments about lack of theory in this area apart from
emerging Industry Convergence.
Risk Capital was consistently viewed as a fundamental ingredient in the industry devel-
opment, but it was not viewed as a primary driver for industrial disruption. Only two
panelists ranked it in the top three, and the few free comments were along lines that for
all types of capital needs, including Radical Innovations or disruptive business models,
there is eventually capital available. However, commenters acknowledged that the right
timing of the access to Risk Capital speeds up disruption.
136
Interrelated economic/ strategic management sub-drivers
All the economic /strategic management disruption sub-drivers were grouped into re-
lated categories the same way as was done with the technology related sub-drivers.
The top eight sub-drivers constituted 96% of the inertia (when requiring that the sub-
driver was present in a minimum of two drivers), and 92% of the acceleration. The ag-
gregated sub-drivers can be seen in Figure 21. The concentration pattern is weaker
than with technological sub-drivers. There are also similarities in technological sub-
drivers and economic/strategic management sub-drivers. The summary of the results
of economic/ strategic management drivers is shown in Figure 22.
Figure 21. Top eight aggregate economic / strategic management sub-drivers
0 20 40 60 80 100 120
Management/ owner beliefs, culture
New business models
Legacy (systems, processes, structure)
Productivity and cost
Industry structure
Platform availability (disipline, industry)
Industry beliefs, culture
Insight
Lack of competences / understanding
Influential early adaptor customers
Focus on continuos improvement
Visionary leaders / experiment culture
Incumbents active resistance
Startups and diginatives
Data (security, quality, availability, ownership)
Growth pressure
Number of points based on the ranking Red = Inertia, Blue = Accelerator
137
Networks, Crowds,
Platforms
Management beliefs Emergence of Platforms
E A U Lack of capabilities Cost benefits (productivity,
capital)
Systems and processes Crowd intelligence effect
Digital Business
Models
Industry barriers Cost benefits (productivity,
capital)
E A U Legacy systems and process-
es
Influential early adaptors
Data (security, ownership) Slow growth of existing busi-
ness
Industry Conver-
gence
Industry beliefs Business model diffusion
E A U Industry-specific financial
steering
Consumer business and be-
havior into B2B
Cultural barriers Generic digital technologies
Figure 22. Summary of the economic / strategic management drivers by the value chain
Note: The colors represent consensus. Green means consensus with the value chain members
who have green color. If one of the three disagrees, red means that the one who disagrees has
stronger support from the green ones and weaker support from the yellow ones. In the driver
section, E = Enabler, A = Actor, and U=User. With sub-drivers, the logic is that the bottom
square is Enabler, the middle square is Actor, and the upper square is User. The inhibitive sub-
drivers are located in the left column and the accelerative sub-drivers in the right column.
138
Results from the survey 4.3
The survey was conducted to investigate five different areas of interest:
1. The systemic impact of the digital technologies and economic concepts (drivers)
on firms’ products, services, operations, and business models (objects)
2. Timing of the potential disruption by the object
3. The disruptiveness of the drivers
4. Preparedness of the respondent’s organization for the potential disruption (in-
vestment)
5. Dependencies in independent variables, drivers, and objects
The same 3x3 background matrix was used in the survey as was used in the Delphi
interviews, and the process was as described in the methodology chapter. The final
number of seminars where the population was sampled was eight (two public, three
invitation-based, and three semipublic, where participation was limited from a popula-
tion that originally was open). The survey was answered by 305 persons. The participa-
tion profiles can be seen in Table 12. The participants came from 23 different countries
and the number of non-Finns was slightly over one-third.
Role/ Value Chain Enablers Actors Users
Scientists 11 10 2
Synthesizers 52 76 31
Decision makers 25 75 23
Table 12. Participation in the survey
Filtering samples with incomplete background information eliminated four samples (one
enabler–decision-maker, one enabler-synthesizer, one actor-synthesizer, and one us-
er-synthesizer). The samples that survived filtering were almost complete; less than 0.5%
of the answers had unanswered data points. In most cases, only one data point was
missing in the sample and therefore the data was completed in order to have other
unique data points from those respondents as part of the analysis.
139
After completion, the data was tested with Shapiro-Wilk test and the results (see Ap-
pendix 5) indicate that the data does not follow a normal distribution. Table 12 reveals
that sub-populations differ from each other, and balancing is required in order to in-
crease the generalization of the results. The value chain is expected to be the main
explanation for the results, and the sample sizes by the value chain vary from 53 to 150.
The scientist sub-populations are so small, even before splitting them based on the
investment behavior, that they are dropped away when analyzing differences in the
value chain with the balanced data. Those samples are not lost, however, as some of
the analysis is done by using the whole population. Leaving out scientists leaves 18
sub-populations for balancing, which results in 278 answers when counting the two
other roles in value chain (see Table 12). Thereafter, 278 is divided by the number of
remaining sub-populations (18), which equals to 15.45 - the total weight of the sub-
population. To get the coefficient for an individual sample, the total weight of the sub-
population is divided by the number of samples in the corresponding sub-population. In
further analysis, selected methods utilize those coefficients (e.g., when calculating
weighted averages or variances).
The last phase before analysis is the interaction effect test using the two-way ANOVA;
the results are enclosed in Appendix 6. The description given by the XLSTAT states
that adjusted R2 express the proportion of the variability of the dependent variable of
the model, and takes into account the number of variables used in the model. The first
goodness of fit is for the whole population (see Appendix 6), and it indicates that the
probability test is not met with Operations, Big Data, and Open Innovation. However, as
only 6.9% and 9.5% of the findings in cases of Operations and Big Data can be ex-
plained, a p-value that is marginally over 0.05 is still acceptable. Open Innovation has
a clearly higher p-value (0.142), but it also has a low explanation power by variance
(8.3 %). The other test (Type III SS) explains how much the combinations of individual
background factors (e.g., place in value chain, role, and investment) can explain vari-
ance in the dependent variable. The test analyzes this with both the combination of two
background factors and with all three. The empty cells indicate that the interactions of
the background variables have 0% chance of explaining the values of the dependent
factor. Type III SS test is recommended if there is an assumption of interaction. The
interpretation by the tool instructions is that lower p-value indicates that the factor has a
larger contribution in the model explaining the variance by the variable examined. The
results suggest that the combination of value chain, role, and investment together
might have interaction in the cases of Big Data, Industrial Internet, and MBSE, but as
the variance can only explain less than 14% of those variables, it cannot be concluded.
140
Results also indicate that individual factors are not very powerful in explaining most of
the variance in data.
Before presenting results, three limitations can be summarized:
Data does not follow normal distribution.
Interaction effect does not shadow the impact of the background factors (inde-
pendent).
Background factors are only capable of explaining a small amount of variance
in the dependent factors (drivers and objects).
The survey results form three categories. First is the whole population that uses unbal-
anced data with an assumption that some of the answers may lie in a larger population.
A second category uses balanced data and is seeking differences and similarities be-
tween the value chain members. The third category analyzes dependencies between
the factors (e.g., place in value chain, investment) and values (e.g., systemic impact or
relative disruptiveness of technologies and concepts), and is mostly impacted by the
limitations above.
4.3.1 Whole population
The first result is the systemic impact of digitalization (i.e., continuous improvement,
quantum leap and disruption). The results of the impact by object can be seen in
Appendix 7. The result is presented by plotting the values with XLSTAT univariate box
plot. In that plot, the box contains 50% of all the values, whereas the vertical line in the
box represents the median value. The lines outside of the box indicate the boundaries,
which include 95% of the entries. If there are still individual points beyond those lines,
they can be considered as improbable outliers. The x represents the average value of
the population. Average and variance values have also been added in the same picture.
The first result indicates the following conclusions:
Data does not follow the normal distribution as the Shapiro-Wilk test indicated;
Product is perceived having least disruption impact (lowest average and vari-
ance);
Digitalization is expected to have quantum leap impact in operations of the firm
that supplies machines (mean and 50% of the answers fit into quantum leap ar-
ea with 3.1 variance);
141
Potential disruption is most likely to take place in the services and business
models (average strong quantum leap but also a high number of answers in dis-
ruption); business models have, however, the highest variance of all of the po-
tential objects; and
Level of variances appears to be high, indicating uncertainty within respondents.
Participants who judged potential disruption in any of the four objects also evaluated
the timing (i.e., how long it takes that the disruption has changed the way of doing
business) in three categories (i.e., less than 5 years, between 5-10 years, or over 10
years). Those results can be seen in Figure 23, which represents the relative shares of
the timing by each object.
Figure 23. Timing of potential disruption by whole population, unbalanced data
The results reveal that although the disruption potential of the products is lowest, the
time to introduce it is the shortest. Services have a high disruption potential and almost
as short an introduction time as products -even faster if the timeline is under 10 years.
Operations have a medium disruption potential and a relatively long lead time, indicat-
ing high inertia to change how a firm operates. Changing business models have a high
disruption potential but the longest lead time to reach acceptance by the industry.
43% 38% 32% 22%
35% 52% 48%
57%
22% 10%
21% 21%
0%
20%
40%
60%
80%
100%
Products Services Operations Business Models
Pe
rce
nta
ge (
%)
Timing of disruption
< 5 years 5-10 years > 10 years
142
4.3.2 By the value chain
The first result by the value chain is the potential disruption impact by digitalization to
products, services, operations, and business models. The used data is now balanced,
but the procedure is the same as with the whole population; the results can be seen in
the plot in Appendix 8.
There are multiple findings drawn from the plot:
The relative position in the disruption (when looking the averages) is the same
with all members of the value chain (i.e., products lowest, operations in the
middle, and services and business models higher);
In all of the objects, users have the highest variance;
Enablers systematically see more disruption potential (average) in all of the ob-
jects than actors and users.
As the potential disruption is hard to judge, and variances are high, the second impact
analysis is based on presenting how the views differ in the value chain with those par-
ticipants who rated the concepts and technologies either weak continuous improve-
ment, or strong disruption. This is based on the assumption that participants who as-
sign strong value farther away from the middle point are more likely have clear argu-
ments for their selections. Those results can be seen in Figure 24.
Figure 24, Weak and strong disruption impacts by the value chain, balanced data
This figure illustrates even more clearly that the disruption expectations are highest in
services and business models, and fair in operations. According to this analysis, the
143
users’ view is clearly more towards disruption in all of the objects. Users also view that
the disruption is not as likely to emerge from the products as enablers and actors be-
lieve.
The disruptiveness of technological and economic drivers was also analyzed by using
XLSTAT univariate box plot; the results are enclosed in Appendix 9. The main observa-
tions from the plot are:
The highest variance is related to 3D printing and Robotics;
Enablers systematically viewed the disruptive impact of the combination of
IoT/Cloud/Big Data to be stronger than did the users and actors;
Enablers are more conservative than users in 3D printing and Robotics;
Economic/ strategic management drivers (e.g., Industrial Internet and Open In-
novation) attract generally higher average values than the technological drivers.
For the same reason as with impact analysis, technology and economic/ strategic
management drivers are also analyzed based on the weak, continuous improvement,
or strong disruption ratings; the results are presented in Figure 25. Also, here the high
variance in 3D printing and Robotics is clearly visible, and users anticipate that these
technologies might have a stronger potential impact than the rest of the population
believes. Open Innovation was identified as an important disruptor by all the value
chain members, but there is a equally sized opposition who perceives very small
impact.
Figure 25. Weak and strong disruption drivers by value chain, balanced data
144
The third result by the values chain is the respondents’ judgments on how well they are
prepared with investments for the potential disruption by digitalization. This analysis
also uses balanced data where the scientists are excluded. The results are presented
in Figure 26. This data clearly indicates that the enablers believe that their company is
taking progressive actions, whereas the users seemed to be equally unprepared, and
the actors land somewhere in between, when almost half of the respondents share the
view that their organizations’ investment in the digital technologies keeps them in the
current relative position compared to their peers.
Figure 26. Impact of own digitalization investments, balanced data
4.3.3 Dependencies
Before analyzing dependencies, various statistical methods were tested, but only corre-
lation and variance analysis appeared to be yielding results. The likely reason for this
might be the high variance in the raw data.
Correlation was analyzed by using non-parametric Kendall’s rank correlation method
for both the balanced and unbalanced data. The Kendall correlation method can be
used for large sample sizes, but Schmidt (1997) pointed out its wide use also in rank
orders. The results for the whole population can be seen in Appendix 10. The results
indicate that majority of the correlations have minor differences in the balanced and
unbalanced data, but they are not significant enough to change conclusions. Therefore,
only balanced data is used when analyzing differences in the value chain, which can be
seen in Appendix 11. The correlation between business model and Industrial Internet is
not taken into account as Industrial Internet is, de facto, a Digital Business Model. The
conclusions from the tables in Appendix 11 are:
25 28 42
34 47 25
41 25 34
0
20
40
60
80
100
Enabler (75) Actor (150) User (53)
Pe
rce
nta
ge (
%)
Regressive Neutral Progressive
145
Values given by actors deliver the weakest correlations between variables;
All participants in the different parts of the value chain view connectivity (IoT) as
a driver for Industrial Internet;
Big Data/AI drives the disruption in all areas in the eyes of enablers;
Big Data/AI, Cloud, IoT form a cluster for Industrial Internet from the view of en-
ablers;
Big Data/AI, Cloud, Iot form a cluster also in the answers given by users;
IoT drives the disruption in all areas in the eyes of users (operations weaker);
Open innovation drives the disruption in all areas from the perspective of actors
and users.
Based on Shapiro-Wilk and univariate box plots, non-parametric methods would have
been more ideal for analyzing variance; however, the use of weighting factors as were
used in the balancing of the sample sizes complicates calculations when using rank-
based comparison methods like Moods median test, as they use unbalanced data by
default. Therefore, ANOVA—which is a parametric method—-shall be applied. Using
ANOVA with non-normally distributed data has a risk of error if the population has a
high number of outliers. This, however, is not as critical an issue as in the correlation
analysis, because ANOVA is using averages of random samples from the sub-
populations, and if the analysis is capable of creating a normal distribution for the sam-
ple averages of the sub-populations, it is also safe to compare the results between sub-
populations. The error is more likely with small sample populations under 30, or if the
variances of the subgroups are very different. As ANOVA is applied in the value chain,
the number of participants by value chain varies between 53 and 150, which reduces
the risk. The variance differences between subgroups are relatively small, as can be
seen in Appendices 8 and 9. The ANOVA F-test determines whether there are differ-
ences between subgroups, and results appear in Table 13. The interpretation of the F-
test is that if the corresponding p-value is less than 0.05 (Nuzzo, 2014), there is a sta-
tistical difference in the opinions between the subgroups (Figures are bolded in the
Table 13).
Pro-
ducts Servi-
ces Oper-ations
Business models 3D
Big Data Cloud IoT II MBSE OI
Robot-ics
0,030 0,099 0,661 0,386 0,004 0,026 0,024 0,003 0,154 0,203 0,448 0,038
Table 13. ANOVA F-test results
146
After the F-test, the data is handled with an ANOVA t-test. The results are shown in
Appendix 12. As the object of the enquiry is to identify differences in the value chain, all
combinations of the value chain where F-test showed less than 0.05 p-value are drawn.
The diagram in Appendix 12 is illustrated with colors, so that if the p-value between two
value chain members is less than 0.05, there is an expected difference in the view. In
those cases, the one in disagreement who believes more in disruption (i.e., higher av-
erage in Appendix 8 or 9) is marked with green, and the one who believes less in dis-
ruption with red. If all disagree, the middle opinion is shaded with yellow (no such cas-
es found). If the t-test did not recognize disagreement between value chain members,
they are marked with blue, indicating that those members of the value chain share the
opinion.
Main findings from Appendix 12 are:
Users are more skeptical towards the claim that products are affected by the
disruption caused by digitalization than are enablers;
Enablers view 3D printing as less disruptive than do actors and users;
Enablers perceive Big Data and Cloud as more disruptive than do actors and
users;
Users are more skeptical towards the impact of the IoT than enablers and ac-
tors are, but believe in Robotization being more disruptive than do enablers.
Cases 4.4
Case descriptions are part of the triangulation. They are, however, building on earlier
cited thoughts of Jouvenel (1967) that the future has to be plausible and case selection
by Eisenhardt and Graebner (1997) that “cases are selected because they are suitable
for illuminating and extending relationships and logic among constructs”.
Two other methods collect data based on the judgment about future development. The
cases illustrate that there is already ongoing disruption in the scope of the research,
even if it is marginal compared to the size of the market from which the cases are taken.
The construct by Eisenhardt and Graebner (1997) is what, in Figure 5, was referred to
as Outcome Economy, or what was modified—based on the Delphi interview round
one in Figure 14—to be more closely suited to the machine-builder. Every case de-
scribes the ultimate output of that business, how the value is created by existing indus-
147
try, and how the new actor has redefined the value creation. In all cases, the value cre-
ation is at higher levels (as seen in Figure 14). The solution has to be based on the
digital technologies within the scope of the research and the application of the technol-
ogies. As one last criterion, the case has to establish a quantum difference in
cost/benefit compared to the current industry practices, and must have originally been
a startup company.
4.4.1 Enevo
Enevo was established in Finland in 2010, using a new business model. They operate
in waste collection (output) with minimal resources in the system. The actors in the
Enevo case are companies that manufacture trucks or waste collection devices at-
tached to the trucks, as well as companies that manufacture recycling and waste con-
tainers. The B2B user is a waste collection company that can be either private or public.
The ultimate cost of waste handling is paid by businesses (cost of doing business) and
citizens (cost of living, or component in taxes if waste collection is a public service).
Enevo is the digital enabler that uses sensors, data transmission, positioning, analyt-
ics/algorithms, and cloud computing as digital technologies.
The waste collection system consists of waste bins and final treatment (e.g., land fill,
recycling, energy), plus transportation and handling in-between. A typical process is
that a truck with waste handling equipment picks up the content of the waste bins and
delivers it for final treatment. The owner has made an agreement with the waste collec-
tion company that, for a certain fee, waste bins are emptied on a regular basis (e.g.,
once a week). Based on these fixed schedules, trucks have their route plans. This pro-
cess results in both half-empty or overfilled bins are picked up, as picking is based on
the calendar rather than on filling patterns.
Enevo’s basic innovation is simple. The containers are emptied only when full, thereby
optimizing the driving time between pickup points. The process is enabled by inexpen-
sive sensors that are attached to the container and powered by a battery that has a
lifetime of up to 10 years in typical conditions and use. The sensors send the infor-
mation on the fill level of the container to Cloud, which-based on time series calculates
the filling pattern and trend of that container, and therefore can predict when it will be
full. That information can be used in route planning. This optimization has multiple logi-
cal values: the need for trucks and waste handling devices is less for every ton of
waste handled. In the same way, the operational cost (e.g., driver, fuel, spare parts)
from moving half empty bins or driving unnecessary distances disappears. According
148
to Enevo, the productivity gain can be as high as 40%; conventional improvement is
following a learning curve of a few percentages per year. The disruption logic is that a
very small investment-in this case, inexpensive sensors and Cloud computing-can en-
able this productivity leap. Also, the improvement of the truck or waste handling equip-
ment is incremental, as the basic solutions have existed for years. Enevo’s solution is
digital, and it can be scaled to fit any city that has a similar waste collection system.
The value creation needs a new business model, or at least a new procurement pro-
cess, where the pickup fee is changed from days of pickups to tons of waste handled.
Using that formula, the service provider can pass part of the savings along to the ulti-
mate payer, and thereby increase its competiveness. The other part of the savings can
be shared between Enevo and the waste collection company via royalties or discounts
of a few cents per ton handled. In the case of recycling, the collecting company sells
the raw material, such as paper, saving in the handling and logistics, thereby yielding
better margins without changing the business model. If one can assume that the need-
ed technology is mature and that everyone in the value chain saves, this should result
in a major growth opportunity. But, there are also losers, and that logically explains part
of the inertia. If the process requires less assets per ton, the need for trucks and waste
handling equipment will go down as long as the entire system reaches a point where
only productive moves are made and all growth of waste translates to the same growth
of needed assets. This is also a difficulty for users who have heavy assets, including a
high number of drivers on the payroll, if they have to compete with new entrants without
assets, and who can apply the new method faster. It is certainly a difficulty for the driv-
ers, as fewer of them are needed to perform the job. If the user is a municipal company,
reducing drivers may be against public opinion. However, in many cases as the amount
of waste is increasing and the fractions collected increase, the cost savings and effi-
ciencies reduce the need to add new resources or invest in new equipment. In addition,
in the private sector, the companies have means to increase their customer bases
without additional investments. As the market is formed by a high number of end-users
(i.e., businesses and citizens), the volume benefit is not realized if only a few of the
end-users join the new system. Logically, one can project a pattern where Enevo’s
growth pattern has a link to Early Adopter logic. When more users (maybe, also, an
Enevo-type competitor) enter the market, at some time there will be a point where us-
ers using the conventional methodology become non-competitive.
Interestingly, in the Enevo case, the actor (i.e., truck manufacturer or waste handling
equipment manufacturer) is not even needed for this productivity leap despite the fact
149
that the innovation reduces the actor’s future market. For truck manufacturers, this
possibility is not even visible in the beginning, as waste disposal is not a major applica-
tion for trucks, and new innovations applied in this segment in a few countries does not
show up in market volumes. However, there might other digital innovations for other
applications where a similar mechanism increases the asset utilization and compound
they might result in lower growth rates that the manufacturer is used to. That still might
be overshadowed by the normal variance of the GDP growth and investment behavior.
For the waste equipment manufacturer, the impact is more direct, but Enevo’s solution
is still outside of market scope and, in the beginning, impact to business volume is
marginal. The reaction of the waste equipment manufacturer depends if they define
their output as waste handling or waste equipment. If the answer is waste handling,
they can try to copy Enevo, join Enevo, or buy Enevo. For Enevo, the strategic ques-
tion is whether Enevo can be become an open platform for all the related players, if the
status of the bin would be a platform that would enable developers, including Enevo, to
develop additional value-adding services, such as recognizing fire before it occurs
based on certain patterns of rising temperature.
The Enevo descriptions are taken from the company’s web page and presentations
that the company has made in public. The case description was read and commented
upon by Mr. Pirkka Palomäki, COO and CTO of Enevo, who verified the logic being
correct.
4.4.2 Trimble
Trimble’s roots stem from 1978 in Silicon Valley based on the navigation and position
technology, and they were an early applier of GPS. For the purpose of this research,
the milestone for Trimble was when they established a joint venture with Caterpillar for
advance guidance and control to be used in agricultural, construction, mining, and
waste industries in 2002. Trimble’s case industry example is farming, as it is also con-
nected to business models that Porter and Heppelman illustrated in their article How
Smart, Connected Products Are Transforming Competition (2014).
In this case, Trimble is the digital enabler, an agricultural farm tractor manufacturer is
the actor, and the farm is the user. The simplified process is that the farm is planting
seeds to be harvested later by using farm harvesting equipment. The output of the pro-
cess is quality and quantity of the crop. For optimal crops, farmers need to know
weather and soil conditions during the growing season as well as the precise applica-
tion rates for chemical and fertilizers based on historical and real-time data. For optimal
150
use of machinery, they need to know the real-time conditions of the machines; if the
farm is large, that need extends to the whole fleet. Farm tractors might have low utiliza-
tion rates outside harvesting season, but during harvest, uptime is mission critical.
Traditional farm output and machinery development has been only partially connected.
The tractor development has followed the logic of any work machine development, per-
formance, features, cost, and regulations that are directed to that particular machine or
application (e.g., safety, noise, emissions). Performance could be improved with higher
horse power, features and innovations (e.g., ability to counter difficult terrains), and
product cost with selection of components and efficiency with the actor’s operations.
Reduction of operational costs, like fuel consumption have been a constant goal for the
product development. The business model for machine-builders has been similar to
that for razor blades, where significant amounts of the profit are made by selling spare
parts during the lifetime of the machine.
Trimble brings the business to Outcome Economy (World Economic Forum, 2015),
goes to higher abstraction level as illustrated in Figure 14, orchestrates the systems of
systems (Porter & Heppelman, 2014), and is building industry-specific platforms that
other players can attach to. They apply sensors, provide data transmission, positioning,
analytics, and cloud computing as digital technologies, combine internal and external
sources of data, and enable the user to monitor and control equipment based on the
knowledge that they have created with digital interfaces like mobile devices. Their
business model is based on software licensing (price per acre per year). They also
offer users additional value-added services that are derived from the knowledge that
they manage the system in the industry. Compared to Enevo, every farm gets the ben-
efit, even if no other farms join the system. In order to obtain industry knowledge,
growth of the number of users is needed.
The same way as in the Enevo example, this all can be achieved without involvement
of the actor as long as the user provides access to the data and Trimble has access
also to resources and knowledge that typically are possessed by the actor. In all in-
stances, better utilization of tractors cuts part of the growth potential of tractors as long
as their systemic development has reached maturity. The tractor manufacturers’ op-
tions are: stay out, develop their own solution, or buy Trimble (which has become less
logical, as Trimble operates in a high number of industries). Caterpillar and Trimble’s
joint venture is closer to acquisition, which has a consequence for Trimble, that in order
to have Connected Farm as an industry platform, the other manufacturers have to co-
operate with Trimble-Caterpillar JV. Compared to Enovo’s example, the business re-
151
definition is already ongoing. Precision farming is close to the concept of Outcome
Economy (World Economic Forum, 2015), and other major actors like Agco27 and John
Deere28 are active in changing the rules of how value is created in farming. The eco-
system around precision farming is not only driven by technological enablers like Trim-
ble or actors like Agco, Caterpillar, or John Deere, but also by companies like Monsan-
to,29 which comes originally from the agricultural chemicals market, but today is invest-
ing hundreds of millions of dollars into what they call Agriculture 2.0, which based on
knowledge from agricultural data.
The data for this description was done by analyzing public sources like the company
web pages, but the logic of the case description was validated by an industry expert of
one of the leading actors in precision farming.
4.4.3 XVELA
XVELA is an internal company startup of the Oakland, California-based software com-
pany, Navis, which was acquired by the Finnish company Cargotec in 2013. Navis
supplies software that enables container terminals to run their operations. XVELA is an
example of an endeavor to build industrial platforms; the development of these plat-
forms came up high in Delphi round two both as an accelerator of Network, Crowds,
Platforms (Figure 22) as well as a generic accelerator of economic concepts that are
sensitive for digitalization (Figure 21).
Navis has been an industry-specific ERP supplier where the value of optimization is
location-specific (i.e., within the confines of the container terminal). Global shipping, as
described in sub-chapter 4.1, is relatively simple in physical form, as global trade is
enabled by goods moving inside of a standardized container. By the same token, how-
ever, the information attached to and enabling the movement of the container and its
contents is highly complex, and there are many stakeholders in every supply chain
transaction (e.g., shippers, forwarding agents, shipping lines, customs, terminals, con-
signees, port authorities, and various governmental departments). The emerging logic
from Delphi round one (Figure 14) was the assumption that many of the disruptive digi-
tal innovations optimize systems at a higher level than actors have been doing. This is
27 http://www.agcocorp.com/precision-farming.html
28 https://www.deere.com/en_US/products/equipment/ag_management_solutions
29 https://ag2point0.com/tag/monsanto/
152
the case also in XVELA, as optimization of a terminal is sub-optimization if you look it
from the supply chain point of view. The amount of waste in global shipping is highly
related to inefficiencies in information flow, and Cargotec has estimated that the lost
value of its waste is $17 billion annually30.
The traditional information flow is based on the exchange of electronic versions of doc-
uments between two trading partners at specific points in time (e.g., after the ship de-
parts a container terminal). For example, a bill of lading is a document that is carrying
the legal ownership from the shipper to consignee, and a shipping manifest contains
the specifics of the cargo. A vessel stowplan document defines all of the containers on
the vessel, their locations, and sufficient information for shipping lines and container
terminals to exchange instructions as to which containers to discharge and load at a
particular terminal. These documents are sent between parties in the supply chain,
including governmental authorities, electronically; they adhere, for the most part to in-
dustry specific standards (e.g. EDIFACT), and are known in the industry as EDI docu-
ments. The crux of the problem is that information is sent to all stakeholders on a 1-to-1
basis and at specific points in time—it is not shared in real time amongst all the parties
who require this information in order to make sound and efficient operational decisions.
Like with the Enevo case, the XVELA basic idea is simple. The shipping information
would be loaded in real time into a Cloud platform that has analytics-based tools for
optimization. Access to information includes security considerations ensuring that all
data is safely shared amongst all the key stakeholders, while safeguarding commercial-
ly competitive aspects of the data. In this regard, security is more robust than what has
existed to date with the sharing of EDI messages among the many parties in the supply
chain. Cost factors that such a solution may impact include:
Capital cost for vessels waiting to be loaded and unloaded;
Extra operational cost (e.g., fuel) for the vessel when sailing at a higher speed
than needed,
Multiple handling of information in the value chain and by other stakeholders,
Lower productivity of the terminal,
Longer storage time for containers in terminals.
30 http://www.cargotec.com/en-global/investors/materials/other-presentations/Pages/CMD-2015.aspx, retrieved 3.5.2016
153
On the revenue side, better knowledge of information would increase vessel utilization
and improve service, enabling performance-based pricing and rewarding good compa-
nies with premiums.
The business model alternatives for XVELA are monthly and usage-based. If Enevo
needs a high number of users at the municipal level to achieve platform benefits,
XVELA needs the same at a global level. Many of the potential users of the systems
are consolidated businesses. If acceleration motives are in productivity and new busi-
ness models, the inertia will likely follow the findings in the Delphi two round. Logically,
a specific issue in this type of collaborative platform is trust (e.g., in Delphi sub-drivers,
trust in a virtual environment), when competitors in the industry load their business data
to the same Cloud platform. The same way as in the Enovo case, better asset utiliza-
tion cuts asset-based business growth (again from actors) somewhat, but actors do not
really have a say whether users join the system or not. All reductions in or automation
of multiple information loading are reductions of the labor force, but the impact is from
so many streams that it is not a major inertia. As major users of the system are large
shipping lines, the inertia results from Delphi two round most likely apply (e.g., man-
agement or industry beliefs, legacy systems), but there are no specific research data
on this. Users in Delphi interviews, however, were aware of XVELA when answering
questions related to Networks, Crowds, Platforms; Cloud, AI, and Digital Business
Models.
The strategic landscape for XVELA does have a few challenges. The first one is the
speed of user acceptance and of reaching a critical mass of participation that will de-
termine whether it can become an industry platform. The other challenge is potential
platform competition. This, according to many of the free comments in Delphi round
two, can lead to a situation where users join several platforms, and none of them
achieve a true network effect. Potential industrial platforms could come from the same
route as XVELA themselves: ERP software competence for ports and terminals. The
other source of potential competition is any player that is in logistics software solutions.
XVELA has a head start, however, and time will show its effectiveness.
This description was created by analyzing public sources, industry (sub-chapter 4.1),
and researcher’s own industry knowledge. XVELA CTO Robert Inchausti also read the
draft and confirmed that disruptive logic of the XVELA case is correctly written.
154
4.4.4 Summary of the three case descriptions
The three case companies all are digital, out-of-industry enablers (from actors point)
whose solutions add major value compared to the historical productivity improvement
of the case industry. The key characteristics of the value-adding output, relationship to
platforms, business models, the use of digital technology, and scalability of the busi-
ness is presented in Table 14, and it will be used as a framework when discussing re-
sults of the two other methodologies in the following chapter.
Enevo Trimble XVELA
Origin (from case
perspective)
New business model
based startup
GPS technology (origi-
nally startup)
Internal startup of an
ERP supplier
Output Cheaper waste per han-
dled ton
Better crop for a farm Transparent, real-time
cargo planning data for
the entire ecosystem
Platform scope Municipal+ Farm+ Global
Business model License fee License fee
(+consulting)
Monthly fee and usage
Digital technology Sensors, wireless data
transmission, Cloud com-
puting, Big Data /AI,
Sensors, wireless data
transmission, Cloud
computing, Big Data /AI,
GPS
Cloud computing, Big
Data / AI
Scalability To similar waste systems To similar farming
methods
Global transportation
Impact to actor Reduced growth rate in
equipment based business
Reduced growth in
equipment based busi-
ness
Chance to leverage
Reduced growth rate of
equipment
Impact to user Reduced cost, opportunity
to new business models
Optimized output Better productivity or
lower cost
Chance to leverage
Table 14. Summary of the three case descriptions
155
5 Discussion and conclusions
The ethos of this dissertation is hermeneutic, and the research question is related to a
complex systemic phenomenon. The intent of the enquiry has been increasing under-
standing of whether the current value creation of the machine-building industry is going
to be disrupted by digitalization and, furthermore, if there is a paradigm change, what
types of technological and economic/ strategic management drivers cause the change.
The underlying assumption has been that the consensus or disagreement about this
future development in the value chain impacts the dynamics, and therefore the value
chain has been a key component in the research design. This systemic complexity of
the question triggered the use of relatively large number of technological and econom-
ic/strategic management perspectives. It also created a practical limit on how detailed
the investigation could be without losing the holistic approach which is a critical attrib-
ute of the entire enquiry. The same reason drove the choice of mixed methods, which
are not only used for triangulation of the results, but also for expanding the breadth that
the different research methods cover. Apart from the three short case descriptions,
expert judgment was at the core of the Delphi interviews as well as the survey.
The logic of the last chapter is that the discussion section covers the three areas of the
enquiry: disruptiveness, value chain, and system perspective. In the first two, the find-
ings of all three methods and the underlying theory are utilized. The systemic section is
based on presenting the key findings in a system dynamic model. The conclusion sec-
tion presents the contribution to industry evolution and strategic management as well
as to practice. The analyses of reliability and validity of the results is done before end-
ing the dissertation to suggest potential future research avenues.
156
Discussion 5.1
5.1.1 Disruption
It is challenging to judge the drivers of disruption without understanding the characteris-
tics of the potential disruption. The proposition was that digital technologies enable
technological and strategic management concepts which can disrupt the machine-
building industry. This dissertation is not trying to prove the proposition, but rather to
increase understanding about the mechanisms that might result from it. Proving the
proposition would have encountered dilemmas in the spirit of the positivistic philosophy
of science: How do we verify with measurement that the disruption has happened?
How do we measure the exact starting point?
Disruption is, in this dissertation, defined as changing a paradigm of the industry. In the
beginning, the definition of paradigm was done by citing Thomas Kuhn (1996), whose
definition originated from scientific paradigms, but it also serves well in the industrial
context:
On the other hand, paradigm stands for the entire constellation of beliefs, values and techniques, and so on shared by the members of given community. On the other hand, it denotes one sort of element in that constellation, the concrete puz-zle solutions which, employed as models or examples, can replace explicit rules as a basis for the remaining puzzles of normal science.
Theories in the second chapter dealt with content of the starting point, such as Domi-
nant Logic (Prahalad and Bettis, 1986), Industry Recipe (Spender, 1989), and Business
Model (Osterwalder, 2004). Other sets of theories looked upon the transformation
mechanisms like Industry Forces (Porter, 1979), S-Curves (Foster, 1987), Technologi-
cal Trajectories (Dosi, 1982), Disruptive Technologies (Bower and Christensen, 1995),
Diffusion of Innovations (Rogers, 2003), and Industry Convergence (Hacklin et al.,
2013). The third set of theories focused on the firms and people in management who
initiate any change at a practical level: strategic schools and tools, Stakeholders
(Freeman, 2015), Prospect Theory (Kahneman and Tversky, 1979), Bounded Ration-
ality (Simon, 1979), and Absorptive Capacity (Cohen and Levinthal, 1990). The frame-
work used to look at the system was based on PEST (Aguilar, 1967). It was originally
developed as a tool to scan business environments, but it took simultaneously into
consideration political, economic, social, and technological factors which were an inte-
gral part of Schumpeter’s (2008) seminal concept of the Creative Destruction, which
also can be labeled changing paradigms.
157
Building on the notion of a heuristic starting point, one can draw from chapter 4.1 a list
of beliefs that are in some form present with the practitioners of global container-
handling.
Users are conservative as they have high amount of capital tied up with the cur-
rent assets which still have a long lifetime;
Users are conservative as many of the them have a structural advantage to be
in a location where high numbers of consumers and suppliers are located;
Actor’s business is driven by trade growth that has a positive outlook, and mar-
ket share of new equipment is a result of competition whose dynamic is best
explained by Porter’s (1979) Five Forces;
The profitability of actors is mainly created with the after-sales services, where
the highest margin component is spare parts;
Unmanned operations are still in their infancy stage, but their emergence is like-
ly to accelerate;
Out-of-industry startups do not possess credibility to seriously enter the busi-
ness.
The disruption itself was empirically investigated in the survey and case descriptions,
and it was also elaborated upon in many of the Delphi round two free comments. For
analytical purposes, the potential impact was analyzed separately by products, ser-
vices, operations, and business models; therefore, the discussion follows the same
logic. It is, however, recognized that concepts like Service of Everything (Zysman et al.,
2015) blur the boundaries of those categories.
Products
The survey result suggests that within the confines of this dissertation, products will not
be the area where disruption emerges by digitalization. The minority, although they
believed in disruption, also believed that the introduction time is shorter than in the dis-
ruption related to the services, operations, or business models. The interpretation is
that both development work and radical innovation of products continue to change
market positions between actors using mechanisms that are well-explained by Porter’s
(1979) Five Forces, but without changing the rules of the value creation of the industry.
Also, new substitute products fit into the same Five Forces or general substitute theo-
ries in economics. There is also history-based evidence, that in global container-
handling, there has been an array of new types of products or methods introduced at
frequent intervals without industrial disruption as defined in this dissertation. Digital
158
technologies are often generic, and more often than not originate from consumer busi-
nesses. Process computers that are needed for controlling the machines are available
for all machine builders serving all industries. This makes it even harder to achieve
product-based sustainable competitive advantage. Applying digital technologies faster
than the competition can create temporal advantages that can be sustained if the actor
is able to repeat its leadership in innovations.
The case descriptions validate the point from a different angle. In the cases where the
disruptive impact was achieved without actors’ involvement and with little or no cou-
pling to the product, even the digital solutions impact the future market of the products.
In the infancy phase of digitalization, machine builders were using own technologies
that kept companies like Trimble away, as machine builders had proprietary compo-
nents that they programmed by themselves. That development did not allow scaling or
external innovation, and it did raise the barriers for the user who wanted to switch to
another supplier. Also, one machine-one industry solutions meant that the R&D in-
vestment did not scale up, and consequently the progress was slow.
The actors’ choices for the product is related to original sensoring (i.e., users can let
sensoring to be installed to existing fleet by someone other than the actors), onboard
computing, technologies that enable actuating functions based on computing, and
technologies that enable data transmission to external systems. The biggest difference
with the actors and companies in the case descriptions is the fact that the data ma-
chines collect from their performance, or from the process of which they are part, or
from the environment where the process takes place, is a vital source of knowledge for
development of new machines, but only for the actor. This does not take away the fact
that in new machine development, boundaries of disruption are at the single machine
level or location-based fleet.
The weakness of product-based industrial disruption can indirectly be derived from the
small number of free comments on the role of the products in the Delphi interviews.
One could summarize them in one comment: Machines have become smarter, and
they will continue to do so.
Services
In the survey data, services were perceived to be slightly more potential object for in-
dustrial disruption than business models (Appendix 7). Figure 23 shows that 90% of the
respondents of the survey, who judged disruption to take place in services, also judged
159
that it will take place within the next 10 years. Predominantly, the “digital disruption”
research reports of machine-building are close to this notion of the GE white paper
(Evans & Annuziata, 2012). This might relate to logic that the objects of the service are
multidisciplinary technological products, and the perspective of disruption is, often, how
one of the existing incumbents could be the disruptor. In business seminars, the same
notion is illustrated with one of the most referred examples: Rolls Royce and power by
the hour.31 It indicates that performance-based service (also defined as a new business
model) is used by Early Adopters (Rogers, 2003); this case suggests that aviation has
been the forerunner at the industry level.
For the digital enabler, it is different, and all the case firms can be defined as service
companies: Enevo in waste management service, Trimble in crop optimization service,
and XVELA in real-time cargo information service. Those services are only weakly
coupled with actors’ products. Actors’ services historically are related to performance of
individual products: spare part supply, maintenance and repairs, and by digitalization-
more towards uptime services that utilize the data that the machine builder has from
the real operation, merged into its design knowledge. Enevo and XVELA services do
not interfere with those types of services. Their services, as well as Trimble’s, is in the
space which is a natural adjacency for actors to grow as firms. Trimble is also in the
actors’ service space, which is enabled by the knowledge of machines they have
through the joint venture with Caterpillar.
As long as the service is attached to the actor’s own equipment, logically the disruption
potential for the industry is limited. Service innovation even with that scope still has
great profit and growth opportunity. The disruption mechanism impacting actors is more
related to whether one player can service different brands, and whether the service is
scalable. Trimble’s is also closer to this development. Trimble is trying to do both, as it
is involved with the machine-building discipline with its connection to Caterpillar, and
also applies the concept not only to agriculture (which was in the case description in
sub-chapter 4.4), but also to other industries like mining and construction. The scalabil-
ity increases with the rate by which value can be added remotely.
The free comments from the Delphi round two were along the same lines, with two ex-
pansions. Service operations are a target of quantum leap by digitalization that will
31 http://www.rolls-royce.com/media/press-releases/yr-2012/121030-the-hour.aspx, retrieved 16.5.2016
160
benefit the user and create competitive advantage for the actor. That builds on logic
that domain knowledge lies with current actors who would not change the content of
the service but just improve the execution. The other comment was also evolutionary,
stating that one avenue to digital services is bundling it with capital. The leasing busi-
ness has already contained various services built-in to the offering. It is not the same
discussion as what was elaborated upon in sharing economy articles, other than the
attitude for ownership might be changing in society at large. Platforms were mentioned
several times in free comments, and most often they were linked to new types of ser-
vices if the user would open its data, but that will be discussed more in the driver sub-
chapter.
Operations
The impact of digitalization on operations is, according to the survey, stronger than on
products but weaker than the impact on service/business models. If the impact is dis-
ruptive, it is perceived to take a longer time than if one initiates the change via a disrup-
tive product innovation. There are a number of interpretations here. On the lower dis-
ruption side, one can argue that software has been an enabler of both continuous im-
provement and quantum leaps for a very long time. This is most evident in process
schools, or re-engineering as Hammer (1990) called it. Processes that were supported
by systems also cause inertia as discussed earlier in the qualitative findings. This most
likely is related to why disruptive transformation was judged to take a long time even
though there were no specific beliefs inhibiting it. Cloud computing that was analyzed
as a driver often impacts operations. As earlier cited, Salesforce.com (Weinhardt et al.,
2009) is a good example of one process as it changes the company-specific ERP solu-
tions into common pay-per-use solutions where the best practice process is incorpo-
rated into the service. From the competitive angle, this commoditizes a key account
process. For the actor, this might be a quantum leap in efficiency of one process, but it
is undeniably disruptive to those whose industry is defined as key account manage-
ment software.
Many digital technologies of enablers allow actors to impact users’ operations. The
case descriptions are such that actors are partially bypassed in the process. Digital
technologies or related management concepts are often available for all value chain
members; Open Innovation can allow very different development processes compared
to closed innovation. These cases do not provide data on disruptiveness to operations
of enablers or actors. However, there is evidence on the disruptive impact to users’
operations. In Enevo’s case, the waste handling company has to change its collection
161
system as directed by analytics in the Cloud platform rather than using fixed routes.
Trimble’s results changes planning on the farm to be based on knowledge of the
weather and other data-driven evidence. If the actor uses Trimble’s service, it can use
the platform to develop services for its users. In that case, the actor uses modules like
positioning, fleet management software, etc. XVELA, for the most part, passes actors
as it is not a value chain-based service, but rather is an omni-channel platform where
all stakeholders can get knowledge to improve their operations. With XVELA, this can
result in major operational changes for the users, like closing departments when data
handling becomes automatic in the Cloud platform.
Delphi interviews were centered on disruption drivers. Keeping this perspective in mind,
AM/3D printing was the only technology which was believed to have power to catalyze
disruptive changes in operations of the actor. The current technological maturity is still
inhibiting this tool, as will be discussed in the drivers’ sub-chapter.
Business models
Business models were ranked as disruptive as services in the survey. The emergence
of the disruption was also judged to take longer than with the three other potential ob-
jects of disruption. Most likely, this is due to the beliefs that are discussed attached to
the inhibitive sub-drivers that are discussed in the next sub-chapter.
All the cases had innovative concepts that had potential as disruptive impacts. Howev-
er, the companies’ own business models were not disruptive. Their solutions enabled
better planning and productivity leaps in operations, but their own business models to
capitalize on these concepts were similar to the B2B software industry: a developer
sells licenses to users. Many software products have become de facto platforms, like
SAP in ERPs or Google in search engines. It appears that disruptive innovation can, at
least in the beginning, be applied with an old business model such as the example of
Google and a newspaper business model (Huberty, 2015). Even actors are bypassed
in Enevo’s and XVELA’s cases; the user can use the solutions to develop new busi-
ness models. Shipping lines can use performance-based pricing enabled by XVELA
the same way as waste-handling companies can pass part of the savings from the
Enevo solution on to its customers with performance-based contracts, and thereby
change the rules of the waste business. Actors can also utilize the Trimble platform and
offer performance-based contracts for users.
162
Free comments from the Delphi interviews supported the nature of the disruption in
business models being close to the view of Outcome Economy (Figure 5) or moving up
in systems (Figure 14), or the Systems of Systems (Porter & Heppelman, 2014). Busi-
ness model innovation (Osterwalder & Pigneur, 2010) itself can be also a small step;
the only element of the Business Model Canvas in Figure 9 that opens the door to Out-
come Economy is the value proposition.
5.1.2 Drivers
What factors, enabled by digitalization, drive disruption in the machine-building industry
serving global container handling?
The empirical input to the question is primarily originating from the results of the Delphi
interview round two. The role of the survey was triangulation of the drivers. Technologi-
cal drivers were the same as in the Delphi interviews, but economic/ strategic man-
agement drivers were reduced to two and simplified to Industrial Internet and Open
Innovation. Reduction was based on a judgment that it would have been hard to judge
seven economic/ strategic management concepts on top of six technological concepts
in a short time. The selected two were based on literature concepts that have specifi-
cally attached to digital technologies. Both technological and economic/ strategic man-
agement drivers were also present in the cases, but in the discussion those will be only
observations.
The logic of discussion is to group drivers as strong, medium, weak, and interrelated
based on the results from the Delphi round two. Results are presented based on Delphi
results, but are discussed using the other methods and literature. Part of the discussion
concerns the sub-drivers that either inhibit or accelerate the impact of the driver.
Strong drivers
The results in sub-chapter 4.2.2 were based on rank ordering of the top three techno-
logical and economic/ strategic management drivers. In both categories, the top rank-
ing was clear, and thereafter the consensus decreased. The strongest technological
driver was Big Data/ Artificial Intelligence and strongest economic/ strategic manage-
ment driver was Networks, Crowds, Platforms. The clear observation is that further
development in both of these areas is beyond the machine-building industry, let alone
individual players. That also suggests that some degree of Convergence of Industries
is ongoing, even though it ranked in the economic drivers. Porter and Heppelman
163
(2014) also highlighted that smart connected products pose a question: “What Busi-
ness am I in?”
Big Data/Artificial Intelligence
Big Data in the context of sense-making has, in this dissertation, been perceived as
one development towards Artificial Intelligence when a vast amount of real-time sensor
data is combined with other types of data to draw relevant business conclusions. That
type of analytics is also a core enabler to Industrial Internet (Evans and Annunziata,
2012), which in this dissertation was later defined more generically as one Digital Busi-
ness Model. The further acceleration of AI is driven by its business opportunities.
Manyika et al. (2013) estimated the potential impact of automation of knowledge work
in 2025 to be between $5.2-6.7 trillion annually. This type of relationship between tech-
nology advancement (push) and immediate market growth (pull) creates trajectories
that impact industry structures, as Dosi (1982) suggested. Big Data/AI is also well-
aligned with what Dussauge et al. (1992) discussed about generic technologies.
Is Big Data/AI also an enabler in strategic management concepts? As it touches the
very principles of knowledge creation, one can easily conclude that it does have an
impact on strategic thinking. I would, however, from the disruption perspective identify
three particular concepts in strategic thinking: Business redefinition, Dominant Logic,
and Core Competence. An argument for business redefinition is that Big Data/AI offers
new sources of knowledge with which to redefine businesses. The concept is partly
related to Design, Cognitive and Learning schools by Minzberg et al. (1988), with a
strong entrepreneurial angle. It has links to earlier scholars’ creative thinking (e.g. Ha-
mel, 1996; Kim and Mauborgne, 2005) in strategy formation, which in today’s language
most often is referred to as business model innovation (Osterwalder and Pigneur,
2010). Business model change was also often mentioned in the free comments of Del-
phi round two as a prerequisite for technology to be disruptive. Link to Dominant Logic
(Prahalad and Bettis, 1986) or Industry Recipes (Spender, 1989) is attached to the
suggestion from Delphi round two comments that all businesses are becoming data-
centric, as enabled by Big Data and advancement in AI. The logic for Core Compe-
tence (Prahalad and Hamel, 1990) is that if the previous thinking is correct—or if it is
correct in machine-building—one could build a competitive advantage using a company
based on core competence in data science, which Evans and Annuziata (2013) sug-
gest as vital future competence. This logic is also similar to the view of Conner and
Prahalad (1996) about the strength of knowledge-based choice versus opportunity-
164
based choice. Big Data/AI, in that context, can be perceived as an amplifier of
knowledge creation.
From this data-centric perspective, it is plausible that inertia for Big Data/AI expansion
is related to data attributes (e.g., confusion and conflicts in data ownership and con-
cerns of data security). Competence is not currently seen as an advantage, but rather
an inhibitor, in building data-centric businesses. The single most important inertia was,
however, the set of management and industry beliefs; there was consensus on that
importance among enablers, actors, and users. Management beliefs are linked in
Chapter two to Prospect theory (Kahneman and Tversky, 1979), Bounded rationality
(Simon, 1979) , and Absorptive Capacity (Cohen and Levinthal, 1990), whereas indus-
try beliefs are more tied with current Dominant Logic (Prahalad and Bettis, 1986) and
Industry recipe (Spender, 1989) . The sources of Dominant Logic (Prahalad and Bettis,
1986) in Figure 8 also explain components that create industry beliefs’ inertia. Trans-
formation to data-centric business is also a large management challenge even if there
would be no specific beliefs inhibiting as suggested by McAfee and Brynjolfsson
(2012).
Sub-drivers that accelerate Big Data/AI were not so much linked to strategic manage-
ment theory; rather, they had more the logic of enabling, benefit, or improvement. The
top five sub-drivers in rank order were: prescriptive potential, knowledge scalability,
process productivity, users opening their data, and advancement in cognitive compu-
ting. Big Data/AI was also the most disruptive driver in the survey, with the second low-
est variance. It also clustered with Cloud and IoT, and correlated with disruption objects
(enablers). All cases had Big Data/AI-based knowledge creation at the core of their
business values.
Networks, Crowds, Platforms
This concept was judged to be the most disruptive digitally enabled economic/strategic
management driver. It relates to network economy theories (including sharing), as well
as to Open Innovation (Chesbrough, 2006) and various platform strategies presented
in Chapter 2. The concepts also are interlinked. When Häcki and Lighton (2001) dis-
cussed the network effect they discussed the importance of platforms even before the
rise of Cloud computing. Sharing is built on using platforms like Uber, but the paradigm
question is also related to ownership (Belk, 2010; Bardhi & Eckhardt, 2012; Nicholas,
2012).
165
Open Innovation (Crowds) and Platforms are interlinked (e.g., in the discussion of in-
termediaries’ roles) (Chesbrough, 2006). The same way to link to AI and Networks is
present in the collective intelligence discussion by Malone et al. (2010). A prominent
finding in the free comments from the Delphi round two was that open systems at large
already gain market share with their own force, and they need platforms; at the same
time, emergence of new platforms will further speed network usage. The other similar
fundamental view was that Five Forces (Porter, 1979) does not well explain the sys-
temic transitions in business and does not recognize platforms as players in that theory.
Just as with Big Data/AI, management beliefs were the major inertia for network econ-
omy to expand. The theories for the beliefs are the same, but the interpretations are
most likely different. Big Data/AI has more understanding and belief questions related
to its importance. Networks, Crowds, Platform also has beliefs that have undesirable
attributes to management, as it challenges structures, culture, ways to operate, and
management capabilities; that was the second-ranked inertia. The third-ranked inhib-
itive sub-drive, current systems and processes, is often the most used as it is based on
facts. If a company has built its processes according to a traditional value chain and
invested into IT systems accordingly, this poses an implementation dilemma even as
the company would like to shift its operating mode towards network style. Emergence
of platforms is the main accelerator of Networks, Crowds, Platforms. A secondary ac-
celerator, flexibility with capacity and lower cost, is a normal, rational argument in any
investment calculation.
Open Innovation was measured in the survey and it ended up higher than most of the
technological drivers in the average rating. It got most very disruptive views (Figure 25)
and at the same also more very weak disrupting views indicating a polarized view of
the population. Open innovation is also perceived as a generic disruptor (impacting
products, services, operations, and business models), the same way as Big Data/AI
based on the correlation analysis.
All case firms had networks and platforms as integral part of their business. Trimble’s
ConnectedFarm concept needs the ecosystem to contribute data for Trimble to provide
better crop output services. Enevo needs users to join its platform to raise the effec-
tiveness of waste management. XVELA is a real-time cargo information platform whose
value is partly based on the Metcalfe’s logic (2013) of how many users the network is
able to attract.
166
Medium drivers
IoT, MBSE, and Robotics were medium technology drivers. IoT was the second-ranked
driver, and was closest to a phenomenon by the view of panelists with logic that device
connectivity is increasingly enabled by multiple drivers like described by Bradley et al.
(2013), and the growth takes place already, independent of individual industry or com-
pany barriers. Why IoT was clearly higher than Cloud may be because Cloud compu-
ting is more independent from firms, whereas in IoT, firms have an important role in
designing the sensoring and connectivity, but even they cannot stop the logarithmic
growth of connected devices. If Big Data/AI can create new knowledge, MBSE was
seen as a tool to model both existing and new knowledge. MBSE was slowed by man-
agement beliefs which were explained by lack of systemic understanding as well as the
benefits of the technology. Legacy systems were factual inertia. As such, MBSE did not
have strong linkage to any theory, but it was rather a rational management choice of
perceived value by investing in a given technology. It appeared that the judgment on
robotics was connected to AI. The importance of robotics appeared to be linked to
whether the panelist judged the intelligence to be in the upper systems or in the robot
itself. If the view was that intelligence was in the upper systems, then the arguments
were built into Big Data/AI and, in the case of robotics, also into machine-learning. The
societal impact of unemployment in autonomous operations was also present as de-
scribed by Frey & Osborne (2013) and Brynjolfsson & McAfee (2012).
IoT was the second most powerful technology driver in the survey. It was also cluster-
ing with Big Data/AI and Cloud. Trimble and Enevo platforms were based on the con-
nected devices but XVELA source was more aligned with the planning systems.
Digital Business Models, Industry Convergence and Radical Innovation were medium
disruptive economic/strategic management drivers. Digital Business Models was close
to Crowds, Networks, Platforms in ranking and also had the same small variance in the
second Delphi interview round. In the free comments, it was obvious that in machine-
building, the most popular Digital Business Model is that introduced by GE as Industrial
Internet (Evans & Annunziata, 2012) which was generally understood by the panelists
as harvesting data from intelligent fleet of machines, processing that data into
knowledge which is developed as a valuable algorithm (productivity, cost saving, safety
etc.), packaging it as a piece of software, and delivering it as a service where the value
is shared between the developer and user. This development was perceived to be in-
hibited by industry barriers, legacy systems, and the data itself (i.e., ownership and
security). Digital Business Model innovation appeared to follow Rogers’s (2003) Diffu-
167
sion of Innovation theory, indicating the importance of use cases. Convergence theory
in Chapter 2 was predominantly discussed as the merger of the ICT industry to other
industries like media or health care. Attributes like clock speed, ecosystems, or diffu-
sion of business models were also present in the free comments and discussion. More
emphasis was, however, given to the convergence of machine-building and service
industries when concepts like Service of Everything (Zysman et al., 2015) combine
physical products, services, and business models, and whose transformation is ICT-
enabled. Radical Innovation was medium high as a disruptive driver. The argumenta-
tion for it was more towards “Black Swan” (as the likelihood of inventing something as
transformational as a container that was described in Chapter 4.1 is unlikely). The en-
deavors for Radical Innovation were often perceived to lead to quantum leap innova-
tions that change market positions between existing players. To accelerate this exper-
iment culture was perceived as a prerequisite.
Industrial Internet was used as a concept to illustrate Digital Business Models in the
survey. It was perceived to be almost as disruptive as Networks, Crowds, Platforms,
which meant it ranked higher than most of the technological concepts. As there were
only two economic/ strategic management concepts in the survey, more conclusions on
the ranking cannot be made. In the case descriptions, it can be argued that Trimble
and Enevo are in the domain of Industrial Internet, whereas XVELA is a marketplace of
knowledge, based on the real-time data.
Weak drivers
Weak drivers did not necessarily have weak correlations to company success, but
based on the Delphi interviews they had weak explanation power as stand-alone to
industrial disruption. According to this logic, Cloud was perceived to be the weakest
technology; based on free comments by the panelists, it has become the de facto
standard of modern computing, but also has been commoditized. If Radical Innovation
was judged as a source of unexpected results, Black Swan’s (Taleb, 2010) AM/3D
printing received similar comments as a technology. This most likely explains why it
ranked as the second weakest disruption driver.
Cloud was also weak in the survey, but it clustered with Big Data/AI and IoT. MBSE
was a medium driver in the Delphi interviews but ended up as weak in the survey as
Cloud. In the Delphi discussions, the panelists appeared to approach MBSE from the
concept of power of modeling, whereas it might be that in the survey the experts were
judging it as a tool.
168
Weak economic drivers of disruption in reverse rank order were: Risk Capital, 6 Forces
(Five Forces added with stakeholders) and Firm, Strategy, Leadership. The panel had
very high consensus that Risk Capital does not cause disruption but is a powerful cata-
lyst in the beginning to speed up disruptions. Six Forces was recognized as one of the
fundamental theories of the industry dynamics as well as a practical tool in portfolio
development. It was ranked sixth out of seven, but with higher variance than Risk Cap-
ital. The free comments highlighted its static nature, lack of platform, and eco-systems.
Firm, Strategy, Management might have been ranked number one if the question
would have been about innovation-based competitive advantage. Visionary leadership
and experiment culture like those described by Ries (2011) were perceived to be tools
to avoid strong industry beliefs or Innovators Dilemma (Christensen, 2011); the same
way as they were prerequisites for Radical Innovation.
The relationship between weak drivers and the three cases is that originally the busi-
nesses had been startups, where experiment culture and visionary leadership of the
firm played important roles in the beginning.
Interrelated sub-drivers
Delphi interviews were the only sources for sub-drivers. Inertia sub-drivers indicated
why panelists do not believe that a technology or economic/strategic management
concept is disruptive or why disruption is slower than it could be or what specifically
slows it down. Accelerative sub-drivers did the opposite. Interrelated sub-driver is sub-
driver that has been suggested to explain behavior of more than one driver. The top
three technological interrelated inertia sub-drivers were: competence/understanding,
data (security, quality, availability, and ownership), and beliefs (management and in-
dustry). Competence and understanding were related to Absorptive Capacity (Cohen
and Levinthal, 1990). Data issues are logical and rational issues to be solved, but data
ownership disputes reflect also a lack of understanding about how different parties cap-
ture the value from more open data-sharing. Various beliefs have already been dis-
cussed as part of theories related to management cognition and behavior.
Productivity, new insight, and new business models were the three most important
technological interrelated sub-drivers that accelerate the digital disruption. Productivity
and new business model are discussed, but insight has not been highlighted yet. In-
sight was discussed in the context of how IoT-based machine, operation, and environ-
ment data creates new engineering insights. Digitalization was also referenced to cre-
ate new insight based on people being connected.
169
Some of the inhibiting sub-drivers of the economic/strategic management drivers were
the same as with the technology. Top three interrelated inhibitive sub-drivers were:
management/owner beliefs and culture, legacy (systems, processes, and structure),
and industry structures. The culture aspect was somewhat stronger to complement the
beliefs, but competence was clearly shared with technological sub-drivers. Industry
structure (e.g., oligopolistic, old stakeholder organizations, established value chain)
was referred to as natural inertia, but also acknowledged as a politically correct excuse
to focus on continuous improvement. Some panelists linked this to Innovator’s Dilem-
ma (Christensen, 2011), as doing what the customers want and doing it a bit better
than last time still pays off.
New business models, productivity/cost, and platforms were perceived as top interre-
lated sub-drivers explaining acceleration of the economic/ strategic management driv-
ers. Overall, it appeared that Facebook, Apple, Uber, and Airbnb have made business
model innovation a popular term. Productivity and cost represented the hard metrics
that management is used to when making investment calculations. If industry struc-
tures were highlighted in inertia, platforms were one of the main topics breaking it. Plat-
form emergence was perceived far more powerful than disruption by startup types of
new entrants. That might also be a chicken or egg question (i.e., emergence of plat-
forms allows startups to join as complementors), as Gawer and Cusamo suggest
(2002).
5.1.3 Value chain
Does the view on drivers differ depending on where the company is in the value chain?
The data for comparison comes from Delphi round two and the survey. The technolog-
ical drivers to compare are: AM/3D printing, Big Data/AI, Cloud, IoT, MBSE, Robotiza-
tion, and economic drivers of Network, Crowds, Platforms and Digital Business Models.
Sub-drivers were presented in the Delphi results, but there is no comparison data from
the survey. The discussion is categorized into consensus or disagreement and is pre-
sented in order of ranking of the Delphi interviews. Context for the discussion is al-
ready-reviewed literature and the container handling industry. One factor that did not
relate to any specific driver was the difference in the preparedness for digital disruption
that was presented in Figure 26, which showed that enablers were best prepared, us-
ers worst, and actors in between.
170
Consensus
Big Data/AI was ranked the top technical driver in the Delphi interview, and tied for the
top spot in the survey results with IoT. The averages were close with all the value chain
members, and the variance was low with the exception of users in the survey. This
result suggests that the machine-building industry serving global container handling is
becoming data-centric, which enables disruption in the industry.
Network, Crowds, Platforms had equal position in the economic/strategic management
drivers. In the Delphi interviews, it was the top economic/strategic management driver,
with low variance and consensus in the value chain. Open innovation had a slightly
lower average than Industrial Internet, but it attracted more than twice the number of
very strong disruption views in the survey. There was a consensus on the position, but
with higher variance than the other four top rankings. The empirical results and earlier
literature on platforms (e.g., Cusamo & Gawer, 2002) or innomediaries (Roijakkers et
al., 2014) suggest that platforms play a critical role in disruption, and they enable the
power of Big Data/AI, using IoT based data or judgement from humans attached to
Open Innovation or Crowd concept.
Internet of Things or connectivity was the second-highest technological driver in the
Delphi interviews, but it had also low variance and consensus in the value chain. The
survey judged its importance even slightly higher, with medium variance and medium
consensus; users rated this slightly lower than enablers and actors. The small differ-
ence with users might explain that users perceive that they might already have some of
the data by old-fashioned means like maintenance records. The conclusion of IoT,
however, appears to be that it creates a new type of real-time data that has not been
available earlier.
Digital Business Models ranked second in the Delphi interviews; it also had consensus
for this position and a low variance. In the survey, it was almost ranked first, with a
consensus in the value chain and low variance apart from the users. The interpretation
with the users might be that their benefit from Industrial Internet is indirect. It depends
on how the actor can translate the new data from IoT to knowledge, and how the bene-
fit is divided. The lack of current supply did show in the accelerative sub-driver, Early
Adopters (Rogers, 2003) with an interpretation that users whose existing purchasing
patterns had been based on ample supply, faced in the introduction phase the opposite
problem, and they might fear not having enough alternatives or being locked in the in-
troduction phase of the new business model. The strong consensus on Digital Busi-
171
ness Models indicates that the value chain is expecting that value will be divided with
the players in the value chain differently than before, and that logic fits into the defini-
tion of a Digital Business Model.
The judgment of Cloud can be summarized that it is the new way of computing, but on
a stand-alone basis—not as an explanation of disruption. This view had a consensus
with the Delphi interviews and the survey.
Disagreement
The disagreement with the value chain was with MBSE, AM/3D printing, and robotics.
In the Delphi interviews, MBSE was held to be the third most important driver for dis-
ruption; there was a consensus on this, but users judged it less critical as seen in Fig-
ure 16. The disagreement was smaller in the survey as seen in Appendix 9. There was,
however, a disagreement between methods: the Delphi panel ranked MBSE clearly
higher than did the experts who participated in the survey. The disagreement in the
value chain is, perhaps, that MBSE as a tool for modeling complex things is more famil-
iar with enablers, for whom it could be a business, and for actors that get a direct bene-
fit from it. The difference in methods might be explained by the fact that in the interview,
the panelists elaborated on the criticality of modeling, but it might be that in the survey,
the tool shadowed the fundamental feature behind it.
Robotization also caused a disagreement in the value chain, but the pattern was the
same by both methods. Enablers did not see it as disruptive, actors perceived it as a
quantum leap, and users saw it as disruptive. The likely explanation for this is that the
enablers are suppliers across the industries, and automatization is nothing new. Actors
might also have supplied automation (e.g., to the processing industry) or even had their
first references to unmanned operations in ports and terminals, and therefore see it as
evolution. Actors seem to acknowledge that when a manned industry transforms to
unmanned, suppliers’ market shares will be divided with a new way. Robotization ap-
pears to be a paradigm change in the eyes of the users. As described in sub-chapter
4.1, the labor in port operations is a central role. It is a powerful stakeholder; the
productivity of the port has a lot to do with the operator’s judgment, and the compe-
tences both in the field and at the office support a manned paradigm.
Additive Manufacturing/3D printing got a weak rating as a disruptive driver. This view
was consistent between the two methodologies; actors and users perceived it higher
than did enablers. The weakness was elaborated on in the Delphi interviews (e.g., the
172
concept is highly disruptive but the technological obstacles are high for the size and
material used in the machine-building). Enablers’ weaker view can be explained that,
for them, it is a promising technology, and if it breaks through they still continue to de-
liver similar services to actors and users who will get the benefit of the technology. Ac-
tors’ perspectives depend on the progress of the technology; if speed, material cover-
age, and cost can be solved, AM/3D printing can change both the design paradigm as
well the production paradigm. Users who rated AM/3D printing highest see both evolu-
tionary and disruptive benefits. How machines that handle containers are manufactured
is not their direct concern.
Printing spare parts would change the dynamics, as argued in the industry sub-chapter;
spare parts are the most profitable single component in the actors’ offering. Global con-
tainer ports and terminals are driven by consumer behavior (i.e., by goods that are in
the containers), and an innovation which could make it possible to traverse the long
distances by Internet in the form of printing code would disrupt transportation, if the
speed of printers would be changed on an industrial scale. The conclusion for AM/3D is
that it was the only technological driver that can become a Black Swan (Taleb, 2010) in
a form that is still unclear today.
5.1.4 The systemic impact
How are disruptive drivers related to each other in the machine-building industry serv-
ing global container handling?
The process evaluated potential disruption drivers separately in the Delphi interviews
and in the survey. The Delphi interviews with iteration ranked sub-drivers that inhibit or
accelerate disruptive technological and economic/strategic management drivers. The
case studies illustrated how there is already ongoing business, within the limits of this
dissertation that contains some of the characteristics of disruption that is being investi-
gated. The strength of mixed methods was applied in the previous sub-chapter where
both consensuses with the methods and the value chain were used when the final
ranking order was discussed.
The whole dissertation has assumed that the studied phenomenon is systemic, which
was the perspective also expressed by many panelists. In order to increase the under-
standing about the systemic impact, the last phase of discussion uses system dynam-
ics. This will be done by applying a Causal Loop Diagram (CLD) (Sterman, 2000). The
CLD is a tool to represent the time delays and causal relationships between taking a
173
decision and its effects on the system. The CLD visualize how causalities and delays
build up the feedbacks and dynamics in a system level.
The illustration is based on the most critical conclusions from the results:
Big Data/AI is the disruptive technology that enables new knowledge creation;
Platforms accelerate the use of external knowledge;
IoT feeds new engineering knowledge to Big Data/AI;
Crowd and Open Innovation enable new judgmental knowledge.
Several causal relationships can be detected between aforementioned drivers. By rep-
resenting these causalities in a single CLD it is seen that causalities form feedback
loops. In the loop the decision is defined as investment in capabilities which were found
critical in the research. The output is new product features, advanced services, effi-
ciency of operations, or new business models. The dynamics between decisions and
its output is affected by delay. In the model, the dependencies that might impact with a
delay are marked with ǁ. The inhibitive or accelerative sub-drivers are added to CLD
from Figures 17 and 21. The model can be seen in Figure 27, which can be defined as
a heuristic system model explaining the characteristics of digitally enabled disruption in
machine-building serving global container handling.
Figure 27. Big Data /AI and Networks, Crowds, Platforms described by CLD
Degree of BigData AI Utilization
Datacompetencies
+
Efficiency ofoperations
Managementbeliefs
Industrybeliefs
Dominantlogic
Pressure toinvest
CapabilityInvestments
Amount ofdata
+Amount of IoT
sensors
Public data
+
Industryconvergence
Amount of relevanthistory data
+
Assets+
Advancement incognitive computing
Use cases
+
+Performingintelligence
New productfeatures
Advancedservices
New bizmodels
Domain intelligenceacquistion
Utilization ofplatforms
Orchestrationcapability
Disciplineintelligenceacquisition
Users opentheir data
Industryactivity
<Managementbeliefs>
Platformcapasity
Companyinteligence
Legacy
Platformsupply
Contributorsupply
Entrepreneurialknowledge workers
174
Figure 27 reveals two loops that impact the firms’ ability to create value. The measura-
ble factors are utilization of Big Data/AI, which here predominantly present the growing
IoT-based data and Platforms that bring the human network effect. Both factors are
slowed down by management beliefs based on the results of the study. Management
beliefs’ concrete impact in this context is investment in capacity, which can be assets
(e.g., licenses, servers, sensoring), and data competencies (e.g., training, recruitment
of data scientists, subcontracting) in an IoT context. In a human network context, the
investment is in orchestration capability based on the results of the study.
In IoT loops, there are delays (marked ǁ) in the loops, as some loops have time ele-
ments that are very difficult to shorten. Adding sensors increases history data very
slowly. But with that logic, adding sensors now is a no-risk decision if believes that data
is the core essence of competitiveness in the future, even you the business case in not
known today. Management beliefs are impacted by the success of own use cases as
well as use cases of own industry and other industries. Based on theory and results of
the dissertation, management beliefs are driven also by other external factors. The
combination of IoT (marked green) and human loops (marked red) drives the value
creation in products, services, operations, and business models. Success has a strong
but delayed impact on management beliefs, meaning that in order to lead the develop-
ment, management has to first act based on a positive belief.
The human loop is more outside firm or industry control. Platform supply has already
emerged, and it is accelerating through new innovation, both in disciplines (e.g., soft-
ware engineering) and industrially (e.g., health care), as well as by multinational tech-
nology companies and startups in narrower spaces. At the same time, platforms attract
small and large incorporated complementors, as well as knowledge workers who wish
to be independent entrepreneurs. This exact impact of this development is more diffi-
cult to forecast than the impact of sensors.
Conclusions 5.2
The dissertation investigated the research question from a hermeneutic perspective in
the philosophy of science, and used mixed methods. The main method was Delphi
interviews. Delphi Round one validated six technological and two economic drivers that
can enable disruption in the machine-building industry, which is serving global contain-
er terminals. It also, together with literature, yielded sub-drivers that either inhibit or
175
accelerate development. The Delphi round two ranked the six technological drivers
together with the original two economic drivers and an additional five economic drivers
that came from the literature and were supported with comments in the Delphi round
one. New sub-drivers were also added in the Delphi round two, and based on the satu-
ration and pattern their significance was analyzed as if they would have been consid-
ered from the beginning. The survey investigated the object and timing of the potential
disruption and triangulated the rank order results of the drivers from the Delphi round
two. The logic of the Delphi and survey methodology was based on the value chain.
The survey also increased understanding of the preparedness of the value chain mem-
bers for the potential disruption. The third method in the enquiry was analysis of three
short case descriptions. Their role was to illustrate that there are already ongoing dis-
ruptions whose mechanisms are under investigation. The critical conclusions were also
illustrated in a heuristic CLD model.
The conclusions are based on the findings where the three methods have found simi-
larities and consensus. As the survey expanded the methodology, some of its strongest
findings are also presented as conclusions if they were supported by the literature. The
conclusions are presented both for their theoretical contributions as well as their rele-
vance to practitioners. In practitioners sub-chapter issues that had disagreement has
more importance as it might indicate industry specific circumstances. The conclusions
in chapters 5.2.1 and 5.2.2 are written in italics.
5.2.1 Industry evolution, strategy formation
Seminal theories within industry dynamics such as Industry Forces (Porter, 1979), S-
Curves (Foster, 1987), Technological Trajectories (Dosi, 1982), Disruptive Technolo-
gies (Bower and Christensen, 1995), and Diffusion of Innovations (Roger, 2003) ex-
plain well how the dynamics work in established industries. Industrial Convergence is a
prominent area that addresses how mechanisms that are familiar in ICT industry dif-
fuse into other industries and sometimes form new industries. Business model innova-
tion, used in this dissertation in the context of Digital Business Models, appears to work
both in industry dynamics and as a factor in industry convergence.
The conclusions based on the empirical findings suggest the following three attributes
to explain the potential disruption.
176
1. New powerful technologies in knowledge creation (Big Data, AI, machine learn-
ing) and unpresented flow of data by connected devices (IoT) enable the move
to higher levels of optimization like those described in the World Economic Fo-
rum about Outcome Economy (2015).
2. Industrial and discipline-based platforms emerge as new disruptive elements in
value creation.
3. Platforms decouple old value chains that transform into networks, which are
joined by startups and crowds.
If the three attributes above are centric in the investigated disruption, then the empirical
findings that inhibit or accelerate the development will have a focal role in strategy for-
mation. Those drivers were similar in technological and economic / strategic manage-
ment drivers, but in order to focus on the most critical ones as conclusions, those that
have same types of meaning are combined. The acceleration in firms is based on in-
vestments, and productivity/cost was number one reason to do that. Investing in digital
disruption led to the new Digital Business Models as the second most important expla-
nation. New insight was the third most important reason (second as to technical drivers
and fourth as to the economic/strategic management drivers), and recognizing the
emergence of platforms was the fourth most powerful explanation (third with economic
drivers and fourth with technological drivers).
Management/owners’ beliefs constituted the most powerful explanation that inhibits the
preparation for the digital disruption. The second most powerful explanation is industry
barriers that include not only the structure of the industry but also the collective beliefs
within the industry. The third strongest explanation was lack of competencies, including
the lack of understanding of the systemic nature of the potential disruption. The two
next most important sub-drivers explaining the inertia were legacy (e.g., systems, pro-
cesses and structure) and data (e.g., ownership and security).
Redefining the language to be compatible with the strategy formation, the earlier re-
sults can be summarized as four statements. Decisions to invest are based on meas-
urable rational economic factors. It has been recognized that digitalization contains
new critical elements in the strategy formation (e.g., Digital Business Models, new in-
sight and Platforms). Decisions not to invest appear to be based more on beliefs (man-
agement or industry) than measurable factors (cost of changing systems, cost of data
security). Lack of understanding the characteristics of the systemic change and lack of
competencies that it requires slows down the transformation.
177
Strategy formation literature was reviewed using the ten strategic schools by Minzberg
et al. (2010). The first conclusion is that any of those strategic schools do not, alone,
offer guidance through potential disruption. Using Minzberg et al.’s (1998) own frame-
work in Appendix 1, the empirical results suggest that the external world in digital dis-
ruption is unpredictable and confusing. Both the findings from the research and
Schumpeterian’s (2008) view of Creative Destruction indicate that in disruption, the
internal processes are natural that reflect larger societal changes than rational busi-
ness planning is able to capture. Based on these assumptions, there are three sug-
gested conclusions for strategy formation:
1. Firms have to be able to mitigate the management beliefs in the firm and indus-
trial level to be able to capture the potential of digital disruption.
It appears that the mechanisms that were present with three interviews in the
introduction (Chapter 1) or examples like Kodak, do repeat themselves. This
conclusion is within the Cognitive School.
2. The change in business environment appears to be systemic and complex, and
therefore strategy formation by learning is effective.
This is close to the Learning School or the ideals of the Lean Startup (Ries,
2011) and it requires either entrepreneurship by a visionary leader and/or a cul-
ture that fosters experimentation.
3. Digital Business Models, when applied with a network model, are a major
change in traditional industries, and they challenge firms’ current capabilities.
This implies that apart from learning, a major competitive advantage can be ob-
tained if, during the transformation, unique capabilities can be created, which is
along the lines of resourced-based theory in strategy formation (Conner & Pra-
halad, 1990).
5.2.2 Practitioners in machine-building industry
The conclusions from a practical perspective are validated in the machine-building in-
dustry that is serving global container operators. It does not exclude learnings that are
applicable also to machine-builders that have other types of users whose business
178
environments have similar characteristics. The previous chapter had conclusions about
the mechanisms of disruption where the value chain members were in consensus.
Those conclusions can well be used also as guidance for practical strategy work. The
specific conclusions here are related to issues of the object of disruption that were
studied only in the survey and elaborated on by the case descriptions, or by such re-
sults that had disagreement in the value chain. Conclusions are written in the perspec-
tive of the actor in italics.
1. Service is the most relevant object of the disruptive impact of digitalization within
the limits of this enquiry.
Service can be viewed from two perspectives: major growth opportunity and source of
competitive advantage (quantum leap) and service that changes the need to be satis-
fied into higher systemic levels as described in the Outcome Economy by World Eco-
nomic Forum (2015), which is a source of industrial disruption.
The first perspective can be categorized in the same way as what GE calls Industrial
Internet (Evans and Annunziata, 2012), which is one Digital Business Model. Based on
the results, users are ready for it but somewhat reluctant to open their data, which is
needed for the business model to emerge. Some users do open their data, which cre-
ates new use cases, which impact management beliefs; then, after the innovator and
Early Adaptor users, the Early Majority also gets interested (Rogers, 2003). The new
business model also opens the door for new entrants like the component suppliers,
system suppliers like ABB, or digital enablers like Trimble to enter the business as
seen in the precision farming example.
The second perspective relates to a new upper level service that makes the whole low-
er level operation or process dramatically more effective. This implies that growth in
services with the new business model is a prerequisite for the incumbents to grow as
companies as the systemic productivity leap in assets reduce the growth rates of ma-
chines as long as a new equilibrium has been reached. All three cases were character-
ized by this perspective.
Compared to literature and findings in the empirical research, it can be argued that
actors’ business environments have been more stable in the past than ongoing digital
transformation enables it to be changed. This is the logical reason why Porter’s Five
Forces (1985) in the industry dynamics, or Positioning and Design School of Minzberg
et al. (1998) in the strategy formation, have had such a central role. Cognitive School is
179
seldom used, as trained professionals are assumed to make rational decisions. In
those industries where the digital disruption has already taken place, the disruptors
seemed to be visionary leaders with entrepreneurial ways of operating, like Steve Jobs,
Mark Zuckerberg, and Elon Musk have demonstrated. This leads to a second practical
conclusion:
2. If the actor wishes to adapt to digital disruption or be a disruptor, it is beneficial
to use learning as a core element in the strategy formation.
Prerequisite for the importance of learning in strategy formation is understanding that
ongoing systemic change is complex and that the current capabilities are not adequate
for winning. It is also builds on a notion that the chances to design or plan are less
effective than experimenting and adjusting. The literature offers multiple choices to
organize disruptive activities related to digitalization, such as stand-alone business
units, centers of excellence, or cross-business-unit steering as suggested by Porter
and Heppelman (2015). The objective of the learning is not only strategy formation, but
also building unique capabilities, which can be translated to the third practical conclu-
sion:
3. Competence in digital technologies with strong machine-building and industry
knowledge can be combined as a core competitive capability.
The essence of the conclusions is grasped by two quotes: “Unlike physical assets,
competencies do not deteriorate as they applied or shared” (Prahalad & Hamel, 1990)
and “The only way to win is to learn faster than anyone else” (Ries, 2011). Based on
the empirical results, orchestration of the network is one of the needed capabilities as
well the leadership skills to lead the transformation; both skills are needed by the in-
cumbents.
MBSE and Robotization were the two drivers where there was a disagreement in the
value chain. Enablers saw both as a business opportunity. Actors saw MBSE as a
powerful thinking tool and a tool to change business, whereas users saw it as an op-
portunity to communicate their needs better to actors. Users saw Robotization as a
change of paradigm in how to run a terminal business, but actors saw it as an evolution
of smart machines. From an actor’s perspective this finding can be translated to action
planning that takes into consideration concerns of both sides.
180
4. Diagnostics on understanding the differences on views of business environment
changes in the value chain can be developed as new kinds of tools for business
planning.
Many of the earlier practical conclusions have the lack of systemic understanding as
one of the root causes and therefore the fifth practical conclusion is:
5. System thinking or CLD as a specific tool has a lot to offer in strategy formation
in machine-building, where it normally is only applied in control systems engi-
neering.
Validity and reliability 5.3
This dissertation has used mixed methods that have been applied conceptually as
seen in Figure 12. All three methods have different issues in their validity and reliability,
and they will be reviewed separately. The Delphi interviews and case descriptions are
grouped under qualitative methods and the survey under quantitative methods. Validity
and reliability will be concluded by considering the use of three separate methods as
one mixed method at the end.
Qualitative methods
Both validity and reliability are originally concepts of positivistic worldview and have an
assumption that a hypothesis can be verified, and the result is not depending on the
researcher (Guba & Lincoln, 1994). In the Delphi interviews, there was an interaction
between the researcher and experts, which implies that the results may differ if the
investigation would have been done by another researcher.
To mitigate this inherent issue between positivism and hermeneutics, several ap-
proaches have been developed. Lincoln and Guba (1985) constructed a concept of
trustworthiness that included Credibility (Is the right thing investigated?), Transferability
(Is right also right in other contexts?), Dependability (In what conditions are findings
consistent?), and Confirmability (How neutral are the results from the researcher?).
Stenbacka (2001) argued that reliability in the traditional sense does not apply in quali-
tative research, but the quality can be judged by validity, generalizability, and careful-
ness as a whole. This requires that informants are correctly chosen, they are able to
181
express their knowledge, and there is a systematic and careful description of the pro-
cess.
The sample strategies have been described previously and the research design and
process descriptions have been integral parts of the methodology in Chapter 3. How-
ever, there are specific quality issues related to the Delphi technique and case studies.
Linstone (2002) summarized the quality issues in eight pitfalls of the Delphi methodolo-
gy, which framework is used below to judge the quality concerns.
1. Discounting the Future
This implies overweighting the short term, an issue also discussed in the management
cognition. The main mitigation for the issue was the design of how panelists were cho-
sen (i.e., three different roles of future) (Kuusi, 1999). It is fair to argue that all three
groups have to deal with the long term, but at least one can assume that scientists are
familiar with the speed at which technology progresses.
2. The Prediction Urge
Humans dislike uncertainty, which leads to prediction even when the expert does not
possess the knowledge on which the judgment should be based. The design of the
dissertation was not trying to achieve a force consensus, thereby easing part of the
pressure and urge to predict. The other quality measure was that order ranking the
disruptive drivers was not possible without argument of inhibitive or accelerative factors
(sub-drivers). That philosophy was derived from the Argumentative Delphi from Kuusi
(1999), but it also served as an indication of expertise.
3. The Simplification Urge
The trap, according to Linstone (2002), is when a reductionist approach is applied to
social/behavioral systems in order to get more simple explanations. This enquiry was
about a complex, systemic phenomenon. Therefore, the PEST approach (Aguilar, 1967)
was used in order to take both technological and economic factors into consideration.
Participants were also using societal and political factors when they judged the devel-
opment of inhibitive or accelerative sub-drivers. The second way to avoid simplification
was the use of free comments in the enquiry. It also served the purpose of acknowl-
edging that the panelists could have views that are not captured by the research design.
The third element of avoiding oversimplification was the construction of the CLD model.
182
It was not, however, tested with a third Delphi round, but in constructing the survey
results and the case descriptions it gave some support.
4. Illusory Expertise
The illusory refers the expertise to forecast outside the own domain expertise. The
means to obtain the relevant expertise relates in this dissertation to panel structure,
and individual members in relation to the research question. One section of the re-
search question relates to the value chain, which implies that there was equal participa-
tion from enablers, actors, and users in the panel. The depth of expertise was as-
sessed by using Kuusi’s (1999) definition on expertise on future: decision-makers, syn-
thesizers, and scientists. The majority of panelist were senior professionals and pos-
sessed true expertise in more than one role, even when they were asked to approach
the questions based on the designated role. This was further utilized in some of the
analysis when the perspective was formed by using the expertise of the group of 36
individuals. The panel was underrepresented with Asians, females, and persons
younger than 30. The open versus closed society has issues related to cultural differ-
ences (Hofstede et al., 2010). The enquiry did not investigate how differences in cultur-
al background impact the view of how digitalization disrupts the machine-building in-
dustry. However, in some individual disruption sub-drivers’ culture was present, and
was not found to be a strong contributor. There were some remarks on national cul-
tures in the free comments.
5. Sloppy Execution
Linstone (2002) refers to sloppy execution, both for the panelist and by the Delphi
manager. Panelist sloppiness relates to Stenbacka’s (2001) question whether the in-
formants are able to express their knowledge properly. Sloppy execution can also re-
late to choosing panels based on friendship or such recommendations that lead to
group thinking. The issue with the Delphi manager relates to expertise in relation to the
panelists, which can impact both what knowledge is shared and the capability to inter-
pret it. The peer relationship most likely is optimal, so that neither one in the interaction
takes on the role of lecturing. The process of searching for the panelists was that the
Delphi manager created a list of potential companies or universities that fit the study’s
purpose, and then searched roles and persons. The selection of decision-makers was
based on titles (e.g., owner, chairman of the board, CEO, director, executive vice pres-
ident) relative to the company where they came from. The introduction was done either
directly or by someone that the Delphi manager knew who was likely to know the per-
183
son identified. The end result was that 28 out of 36 of the panelists in the Delphi round
two had no prior relationship to the Delphi manager. The logic of the interview was also
based on the time being spent on those drivers that the panelist judged relevant as
described both in Chapters 3 and 4. There were also a couple of test interviews prior to
the Delphi rounds in order to refine the process. The Delphi manager (i.e. the re-
searcher of the dissertation) was assumed -based on his education and industry back-
ground- to be qualified to conduct the interview as well to interpret the free comments,
but obviously it is a biased judgment. The other interpretation issue was not recording
the interviews, which was a tradeoff decision as described in the methodology sub-
chapter 3.1.2.
In the Delphi round two, there was a qualitative self-judgment measure after each in-
terview. In the self-judgment, the session was ranked with four parameters and scale of
one to five, where five was an acknowledged international expert and one was some-
one not adequate for the purpose of the research. The parameters assessed were ad-
equacy of the time, substance knowledge (technology and economic/ strategic man-
agement), ability to apply the expertise to the research question, and the average of
first three parameters, which was a quality measurement of the interview based on how
the researcher judged the interview’s success. In order to count in the results, the
quality needed to be at least 3 out of 5 points; otherwise, the interview would be con-
tinued. Two interviews were continued within a week from interruption, and one was
replaced by a new interview with an expert with the same position in 3x3 matrix.
6. Optimism – Pessimism Bias
This trap is related to the earlier comment of over-optimism in the short term and over-
pessimism in the long term. The other perspective of Linstone (2002) is that optimism
and pessimism are also personal character traits. Availability Heuristics (Kahneman,
2003) also is an issue (e.g., management consulting companies have actively pub-
lished reports on how digitalization is transforming businesses, which is likely to impact
the views). The survey is not mitigating this, apart from being a larger sample than the
Delphi interviews. The analysis of the sub-drivers that inhibit or accelerate development
reduces some of the concern. Also, the objective of the system model was to illustrate
the interrelation of the drivers and sub-drivers, which takes another perspective of the
likely development.
184
7. Overselling
There are a few traps included in the overselling, over-usage of the method for the
same question (the same panelists), anonymity, forced consensus, or hidden agendas.
The researcher was not aware that any of the panelists had previously participated in a
Delphi panel that was investigating the same research problem. Also, as referenced
earlier, there was no forced consensus in the process. Anonymity was maintained dur-
ing the interviews, but the participants were aware of what type of companies or uni-
versities were in the panel and what types of experts the other panelists were. The hid-
den agenda suggests that there would be a vested interest either by the panelist or the
Delphi manager to get a certain type of result. This issue most probably was most
problematic with enablers, as they were also suppliers to actors and enablers. However,
their coverage in digital technologies was not based on a single technology such as 3D
printing, and therefore it is fair to assume that they did not push any particular conclu-
sion to be formed. For the researcher, the intent was the right answer, as false but
pleasant results would logically not lead to a useful strategy for the background organi-
zation of the researcher.
8. Deception
Deception is related both to manipulation and integrity. Manipulation can be regarded
as the more severe form of hidden agenda that was reviewed in the previous para-
graph. A researcher cannot objectively judge his or her own honesty apart from the
logic that there is no motive to promote a specific picture of future development other
than being as truthful as possible.
As described in the motivation (sub-chapter 1.1) the Delphi manager is himself a ma-
chine-builder and consequently has a dual role as a researcher and stakeholder. From
validity perspective this poses an objectivity issue for Information Policy (Kuusi 1999) of
the interviewees and for interpreting the results by the Delphi manager. The magnitude
of drivers, sub-drivers, their abstraction level, industry level perspective and time hori-
zon that is beyond ordinary business transactions most likely has mitigated partly the
issue of objectivity.
Case descriptions
The case descriptions were brief and not strong enough to form an individual research
method. Their application was to describe, from adjacent industries, examples that
possess disruptive features of digitalization according to the literature. Cases were tak-
185
en from public sources, but they were validated by experts who knew the cases, indus-
tries, and the uses of digital technologies. Suitability of the cases were related to Out-
come Economy (World Economic Forum 2015), the preliminary finding on systemic
characteristics of disruption (Figure 14) from the Delphi round one, and the concept of
Systems of Systems by Porter and Heppelman (2015). Therefore, it can be argued that
the issues of validity and reliability have connection to theory building from the cases.
Eisenhardt and Graebner (2007) were discussing part of the problem with the thought:
“Better stories versus better theories.” They opinioned that particularly with thin stories,
like the short descriptions in this dissertation, tables are practical tools to highlight the
essence of the constructs. Table 14 where the cases were summarized was an attempt
to do that.
Quantitative methods
The use of quantitative methods in triangulation and expansion does not take the ethos
of the research away from Hermeneutics. There are, however, different types of validity
and reliability issues in quantitative methods. Reliability, according to Olkkonen (1994)
is particularly relevant in statistical analysis, as statistical criteria can tell the probability
that the results are reliable and indirectly also if the same result can be achieved by
repeating the study. The statistical criterion was integrated in the results in sub-chapter
4.3. Repeatability of the results is still most likely hard to obtain for two reasons con-
cerning temporality. First, the publicity of digitalization appears to have grown since the
experiment was done, which most likely will have an impact on the judgment of poten-
tial participants. The second, more factual reason is that the related technology ad-
vances quickly, as well as does the number of cases applying it.
Validity in quantitative research according to Creswell (2013) is whether one can draw
meaningful and useful inferences from scores on the instruments, which can be further
split into:
Content Validity (Do the items measure the content they were intended to
measure?)
Predictive or Concurrent Validity (Do scores predict a criterion measure? Or
do results correlate with other results?)
Construct Validity (Do items measure hypothetical constructs or concepts?)
The basic questions in the survey were simple in their nature. How strong is an impact,
to an object, and how strong are drivers? Logically, the main Content Validity concern
186
was whether the respondents understood the meaning of the objects and drivers, and if
they understood them in the same way.
The format of the process, which included introduction of the concepts, was chosen to
mitigate the issue of respondents’ understanding of the variables. The formatted intro-
duction created a burden to the researcher, who had to conduct the introduction every
time in the same way. To accomplish this, the researcher created a script specifying
how the concepts were introduced orally apart from the text that the audience saw, in
order to improve the consistency.
Predictive Validity appeared to have some issues beyond triangulation. The sample
population using the 3x3 matrix should have been quite a bit larger than 305, in order
to increase the potential to generalize the results (e.g., to other industries). High vari-
ance in the raw data might also have originated from the fact that it is truly difficult to
judge the systemic impact of digitalization, and there is confusion on the issue. Concur-
rent validity was not possible to check, as the research was not using a method that
would have applied in comparable research questions including the same criterion.
Was the construct valid (i.e., were the drivers in the survey the right ones for measuring
digital disruption)? The process of picking the original drivers was based on recent re-
search reports of commercial research institutes that either specialized in digital tech-
nologies (e.g., Gartner, Cisco, Accenture and IBM) or more generic institutions that
were also researching digitalization from a perspective of strategic management (e.g.,
McKinsey & Co). Delphi round one interviews also served the purpose of validating
this.
Mixed methods
The use of mixed design was based on an assumption that disruption enabled by digi-
tal technologies or economic concepts is a systemic phenomenon, and it is difficult to
form full understanding of the research question. The use of several methods expand-
ed the coverage as illustrated in Figure 12. In the same figure, one can see that the
Delphi one interview validated also the survey (drivers) design, validated Delphi round
two drivers, and suggested sub-drivers as well. Delphi round one results were also
used in the case design.
187
Results were presented as stand-alone as well as the judgment on validity. In the dis-
cussion and conclusion phases, the second strength of mixed methods came through
triangulation. Both the survey results and case description were used to validate find-
ings from the Delphi round two interviews.
Future research avenues 5.4
The components in this dissertation related to industry evolution, strategy formation,
and digitalization as a systemic phenomenon that could have disruptive impact on the
machine-building industry serving global container terminals. The research design uti-
lized mixed methods. It also appears that development of digital technologies and their
applications in business is moving quickly. These premises open the first two natural
research avenues: (1) First, to repeat the study with the methodology in an adjacent
industry that meets the same criterion as described in the sub-chapter 1.2. This avenue
would create comparison data to draw conclusion whether the disruption and its mech-
anisms have similarities between industries. (2) Second, another plausible way is us-
ing exactly the same design; and repeating the study in the global container handling
industry. The purpose of that alternative would be justified with the assumption that the
topic is moving quickly, and that some of the conclusions of this study could be tem-
poral.
Different perspectives of future research could be derived from the main conclusions:
Big Data/AI and Networks, Crowds, Platforms. The logical direction would be to delve
deeper into the mechanisms. It could be done either by continuing the study in global
container handling or by constructing a multi-industry design. One attractive method
could be a case study with an objective to construct a more detailed framework of the
disruption mechanisms. The other alternative of building from the conclusions would
be an extensive literature review of how the conclusion would be able to contribute
generically to the gap that industry dynamics by Five Forces or various innovation
based trajectories have when platforms and ecosystems emerge into capital intensive
industries. Deeper or wider approaches based on the conclusion would work also in the
strategy formation. A theoretical approach there could be an analysis of the impact of
platforms and open systems on existing strategy schools.
The third set of further research avenues relate to the methodology. This dissertation
did find Kuusi’s Argumentative Delphi useful when it was applied with rank ordering
188
together with inhibitive and accelerative sub-drivers that impacted future development.
This was triangulated with a survey. An interesting alternative would be developing the
methodology into online Delphi, where those two methods would be merged. That
would require special attention to user interface in order to motivate the panelists. That
could also incorporate use of visual illustrations (Adelson & Aroni, 2002) to decrease
the variance in understanding the arguments. Pre-programmed analytics could further
enable instant feedback to the panel, maybe even shown as a system dynamic model.
The use of scenarios based on the Delphi findings as suggested by Mannermaa (1999)
or applying the futures Map (Kuusi et al. 2015) could be methodologically also an inter-
esting direction to continue the research. One of the most suitable technologies as a
research object is Additive Manufacturing / 3D printing as it got quite polarized re-
sponses, and also got several suggestions of being a wild card or Black Swan.
The last set of research possibilities centers around management. If the conclusions
are correct, what does it imply in day-to-day strategic management? Does the firm truly
have a chance to disrupt the industry, or is the development such that it is beyond the
incumbents of traditional industries? Even so, based on the findings of this research,
business model innovation and experimental culture are advantageous in achieving
quantum leaps in competiveness. Also, a practical checklist of choices, like the one
Porter and Heppelman (2014) suggest, are useful, and one could study the conditions
that impact how to make those choices.
If the assumption in the beginning of this dissertation is true -that the ability to trans-
form any kind of data, observations, events and so on to a digital form, by processing it,
storing it, and sharing it, economically, globally, and in real time -has the same kind of
systemic characteristics as the invention of writing or printing, the theme will stay a rich
source of further research for years to come.
189
References
Abernathy, WJ and Utterback JM (1978). Patterns of Industrial Innovation, Technology
Review, Vol 80, Issue 7, pp. 41-47
Achrol, Ravi S.; Kotler Philip (1999). Marketing in the Network Economy, Journal of
Marketing, Vol. 63, Issue 4, pp. 146-163
Adelson, Marvin; Aroni, Samuel (2002). Differential Images of the Future, The Delphi
Method : Techniques and Applications, Edited by Linstone, Harold A.; Turoff Murray,
academia.edu, pp. 430-432
Afuah, Allan N; Utterback, James M. (1997). Responding to Structural Industry
Changes: A Technological Evolution Perspective, Industrial and Corporate Change, Vol
6, Issue 1 pp. 183-202
Aguilar, Francis Joseph (1967). Scanning the Business Enviroment, Collier-Macmillan,
Toronto, Canada, pp. 1-17
Alphaliner (2016). Monthly monitor, www. alphaliner.com, May, p. 12
Ansoff Igor H. (1957). Strategies for Diversification, Harvard Business Review, Vol. 35,
Issue 5, pp.113-124
Arora, Ashish; Fosfuri, Andrea; Gambardella, Alfonso (2001). Markets for Technology
and Their Implications for Corporate Strategy, Industrial & Corpote Change, Vol. 10,
Issue 2, pp. 419-451
Bardhi, Fleura; Eckhardt, Giana M. (2012). Access-Based Consumption: The Case of
Car Sharing, Journal of Consumer Research, Vol. 39, Issue 4, pp. 881-898
Barney, Jay B. (1986). Organizational Culture: Can It Be a Source of Sustained
Competitie Advantage?, Academy of Management Review, Vol. 11. Issue 3, pp. 656-
665
Baruch, Yehuda (1999). Response Rate in Academic Studies – A Comparative
Analysis, Human Relations, Vol. 52, Issue 4, pp. 421-438
190
Beer, Michael; Eisenstat, Russel A.; Spector, Bert (1990). Why Change Programs
Don’t Produce Change, Harvard Business Review, Vol. 68, Issue 6. pp. 158-166
Belk, Russel, (2010). Sharing, Journal of Consumer Reasearch, Vol. 36, Issue 5,
pp.715-734
Benbasat, Izak; Goldstein, David K.; Mead, Melissa (1987). The Case Research
Strategy in Studies of Information Systems, MIS Quarterly, Vol. 11, Issue 3, pp. 369-
386
Benner, Mary J.; Ranganathan, Ram (2013), Divergent Reactions to Convergent
Strategies: Investor Beliefs and Analyst Reactions During Technological Change,
Organizational Science, Vol. 24, Issue 2, pp.378-394
Bergek, Anna; Berggren, Christian; Magnusson, Thomas; Hobday, Michael (2013).
Technological Discontinuities and the Challenge for Incumbent Firms: Destrcution,
Disruption or Creative Accumulation?, Research Policy, Vol. 42, Issue 6-7, pp.1210-
1224
Betker, Michael; Fernando, John; Whalen, Shaun. (1997). The History of
Microprocessor, Bell Labs Technical Journal, Vol. 2, Issue 4, pp. 29-56
Bower, Joseph L.; Christensen, Clayton M. (1995). Disruptive Technologies: Catching
the Wave, Harvard Business Review, Vol 73, Issue 1, pp. 43-53
Bradley, Joseph; Barber, Joel; Handler, Doug (2013). Embracing the Internet of
Everything to Capture Your Share of $ 14.4 Trillion,
http://www.cisco.com/web/about/ac79/docs/innov/IoE_Economy.pdf, p.1. retrieved,
12.8.2015
Brown, Brad; Court, David; McGuire, Tim (2014). Views from the Front Lines of the
Data-Analytics Revolution, McKinsey Quarterly, Issue 1, pp. 94-101
Brynjolfsson, Eric; Hu, “Jeffrey” Yu; Smith, Michael D. (2006). From Niches to Riches:
Anatomy of the Long Tail, MIT Sloan Management Review, Vol. 47, Issue 4, pp. 67-71
Brynjolfsson, Eric; McAfee, Andrew (2012). Thriving in the Automated Economy, The
Futurist, Vol. 46, Issue 2, pp. 27-31
191
Burt G.; Wright G.; Bradfield R.; Cairns G.; van der Heijden K. (2006). The Role of
Scenario Planning in Exploring the Environment in View of the Limitations of PEST and
Its Derivates, International Studies of Management & Organisations, Vol. 36, Issue 3,
pp. 50-76
Cable, Daniel M; Shane, Scott (1997). A Prisoner’s Dilemma Approach to Entrpreneur-
Venture Capitalist Relationships, Academy of Management Review, Vol. 22, Issue 1,
pp. 142-178
Chandler, Alfred D. Jr. (1990 3rd ed.). Strategy and Structure: Chapters in the History of
the Industrial Enterprise, MIT Press, Cambridge, USA, pp. 14-17
Charitou, Constantinos D.; Markides, Constantinos C. (2003). Responses to Disrruptive
Strategic Innovation, MIT Sloan Management Review, Vol. 44, Issue 2, pp 55-63
Chesbrough, Henry W. (2003). The Era of Open Innovation, MIT Sloan Management
Review, Vol. 44, Issue 3, pp. 35-41
Chesbrough, Henry (2006). Open Innovation: A New Paradigm for Understanding
Industrial Innovation. Open Innovation: Researching a New Paradigm, Edited by Henry
Chesbrough, Wim Vanhaverbeke and Joel West, Oxford University Press, Oxford , UK,
p. 1
Chesbrough, Henry (2010). Business Model Innovation: Opportunities and Barriers,
Long Range Planning, Vol. 43, Issue 2-3, pp. 354-363
Chesbrough, Henry; Ghafele Roya (2014). Open Innovation and Intellectual Property,
New Frontiers in Open Innovation, Edited by Henry Chesbrough, Wim Vanhaverbeke
and Joel West, Oxford University Press, Oxford, UK, pp. 192-207
Child, John (1972). Organization Structure, Environment and Performance: The Role of
Strategic Choice, Sociology, Vol 6, Issue 1, pp. 1-22
Childress, Diana (2008). Johannes Gutenberg and the Printing Press, Twenty-First
Century Books, Minneapolis, USA, pp. 59-61
Christensen, Clayton M. (2011 4th ed.). The Innovators Dilemma, HarperCollins, New York, USA, pp. xix, 257
192
Christensen, Clayton M.; Johnson, Mark W.; Rigby, Darrel, K. (2002). Foundations for
Growth, How to Identify and Build Disruptive New Businesses, MIT Sloan Management
Review, Vol. 43, Issue 3, pp. 22-31
Christensen Jens, Froslev; Olesen Michael, Holm; Kjaer Jonas, Sorth (2005). The In-
dustrial Dynamics of Open Innovation – Evidence from Transformation of Consumer
Electronics, Research Policy, Vol 34, Issue 10, pp. 1533-1549
Clemmons, Paul, Geller, Jason, Messe,r Adam (2012). Cloud Hits the Enterprise,
Deloitte Review, Issue 11 p.46
Cohen, Daniel; Sargeant, Matthew; Somers, Ken (2014). 3-D Printing Takes Shape,
McKinsey Quartely, Issue 1, pp. 40-45
Cohen, Wesley M.; Levinthal, Daniel A. (1990). Absorptive Capacity: A New
Perspective on Learning and Innovation, Administrative Science Quarterly, Vol. 35,
Issue 1, pp. 128-152
Collis, David J.; Montgomery, Cynthia A. (1998). Creating Corporate Advantage,
Harvard Business Review, Vol. 76, Issue 3, pp. 70-83
Conner, Kathleen R. (1991). A Historical Comparison of Resource Based Theory and
Five Schools of Thought Within Industrial Organization Economics: Do We Have a New
Theory of the Firm?, Journal of Management, Vol. 17, Issue.1, pp 121-154
Conner, Kathleen R.; Prahalad C.K. (1996). A Resource-based Theory of the Firm:
Knowledge Versus Opportunism, Organization Science, Vol. 7, Issue 5, pp. 477-501
Cooper, Robert G. (1990). State-Gate Systems: A New Tool for Managing New
Products, Business Horizons, Vol. 33, Issue 3, pp.44-54
Cotteleer, Mark, J. (2014). 3D Opportunity for Production, Deloitte Review, Issue 15, pp.
146-161
Creswell, John, W. (2003). Research Design, Qualitative, Quantitative and Mixed
Methods, Saga Puplications, London, UK, pp. 6,15,16,160
Cummins, Stephen, Wilson David, 2003. Images of Strategy, Blackwell Publishing,
Oxford, UK, p. v
193
Curran, Clive-Steven; Bröring, Stefanie; Leker, Jens (2010). Anticipating Converging
Industries Using Publicly Available Data, Technological Forecasting and Social Change,
Vol. 77, Issue 3, pp. 385-395
Dalkey, Norman; Helmer, Olaf (1963). An Experimental Application of the Delphi
Method to the Use of Experts, Management Science, Vol. 9, Issue 3, pp. 458-467
Daniel, Ronald D. (1961). Management Information Crisis, Harvard Business Review,
Vol. 39, Issue 5, pp. 111-121
Danneels, Erwin (2004). Disruptive Technology Reconsided: A Critique and Research
Agenda, Journal of Product Innovation Management, Vol. 21, Issue 4, pp. 246-258
D’Aveni, Richard (2015). The 3-D Printing Revolution, Harvard Business Review, Vol.
93, Issue 5, pp. 40-48
Dawson, Jeremy F. (2014). Moderation in Management Research: What, Why, When
and How, Journal of Business & Psychology, Vol. 29, Issue 1, pp. 1-19
Dedehayir, Ozgur; Mäkinen, Saku J. (2011). Measuring Industry Clockspeed in the
Systemic Industry Context, Technovation, Vol. 31, Issue 12, pp. 627-637
De Jouvenel, Bertrand (1967). The Art of Conjecture, Basic Books, New York, USA, p.
18
Deschamps, Jean-Philippe (2005). Different Leadership Skills for Different Innovation
Strategies, Strategy and Leadership, Vol. 33, Issue 5, pp. 31-38
Dosi, Giovanni (1982). Technological Paradigms and Technological Trajectories: A
Suggested Interpretation of the Determinants and Directions of Technical Change,
Research Policy, Vol 11, Issue 3, pp.147-162
Doz, Yves L..; Kosonen, Mikko (2010). Embedding Strategic Agility; A Leadeship
Agenda for Accelerating Business Model Renewal, Long Range Planning, Vol. 43,
Issue 2-3, pp. 370-382
Drewry Maritime Research (2014). Global Container Terminal Operators Report, pp.3,
12,16
194
Drucker, Peter F. Managing in the Next Society (2002), Truman Talley Books, New
York, USA, pp. 22-24
Dussage, Pierre; Hart, Stuart; Ramanantsoa, Bernard (1992). Strategic Technology
Management, John Wiley & Sons, New York, USA, pp. 106-108
Early, James M. (2001). Out to Murray Hill to Play: An Early History of Transistors,
IEEE Transactions in Electron Devices, Vol. 48, Issue 11, p. 2468-2472
Ebert, Christof (2008). A Brief History of Software Technology, IEEE Software, Vol 25,
Issue 6, pp. 22-25
Eisenhardt, Kathleen; Graebner, Melissa (2007). Theory Building from Cases:
Opportunities and Challenges, Academy of Management Journal, Vol. 50, Issue 1, pp.
25-32
Evans, Peter C.; Annunziata, Marco (2012). Industrial Internet: Pushing the Boundries
of Minds and Machines, General Electric White Paper, www.ge.com, retrieved
27.6.2014, p. 4
Fabbris, Luigi (2013). Measurement Scales for Scoring or Ranking Sets of Interrelated
Items, Survey Data Collection and Integration, Edited by Davino, Cristina; Fabbris,
Luigi, Springer Verlag, Berlin / Heidelberg, Germany, pp. 21-43
Fai, Felicia; von Tunzelmann, Nicholas (2001). Industry-specific Competencies and
Converging Technological Systems: Evidence from Patents, Structural Changes and
Economic Dynamics, Vol. 12, Issue 2, pp. 141-170
Felin, Teppo; Zenger, Todd R. (2014). Closed or Open Innovation? Problem Solving
and the Governance, Research Policy, Vol. 43, Issue 5, pp. 914-925
Fischer, Roger S. (2008). A History of Writing, Reaktion Books Ltd, London, UK, pp.
27-33
Forrester, Jay (1994). System Dynamics, Systems Thinking, and Soft OR, System
Dynamic Review, Vol. 10, Issue 2-3, pp. 245-256
Foster, Richard N. (1986). Innovation: The Attacker’s Advantage, Summit Books, New
York, USA, p. 101-106
195
Freeman, Edward, R. (2015). Strategic Management: A Stakeholder Approach,
Cambridge University Press, Cambridge, UK, p. 25, 141
Frey, Carl Benedikt; Osborne, Michael A. (2013). The Future of Employment: How
Suspectible Are Jobs to Computerisation?, University of Oxford, UK, pp. 36-42
Gadamer, Hans-Georg (2013). Truths and Methods, Bloomsbury Publishing, New York,
USA, pp. 3-9, 378
Gambardella, Alfonso; Torrisi, Salvatore (1998). Does Technological Convergence
Imply Convergence in Markets: Evidence from the Electronics Industry, Research
Policy, Vol 27, Issue 5, pp. 445-463
Garetti, Marco; Rosa, Paolo; Terzi, Sergio (2012). Life Cycle Simulation for the Design
of Product-Service Systems, Computers in Industry, Vol. 63, Issue 4, pp. 361-369
Garland, Ron (1991). The Mid-Point on a Rating Scale: Is It Desirable, Marketing
Bulletin, Vol 2, Issue 2, pp. 66-70
Garrison, Gary (2009). As Assessment of Organizational Size and Sense Response
Capability on the Early Adaption of Disruptive Technology, Computers in Human
Behavior, Vol. 25, Issue 2, pp.444-449
Gawer, Annabelle; Cusamo, Michael A. (2002). Platform Leadership, How Intel,
Microsoft and Cisco Drive Industry Innovation, Harvard Business School Press, Boston,
USA, pp. 8-10
Gawer, Annabelle; Cusamo, Michael A. (2008). How Companies Become Platform
Leaders, MIT Sloan Management Review, Vol. 49, Issue 2, pp. 28-35
Gerstner Louis Jr. (2002). Who Says Elephants Can’t Dance, HarperCollins Publishers,
New York, USA
Godet, Michael (1994). From Anticipation to Action, UNESCO Publishing, Paris, France,
p. 29-32
Govindrajan, Vijay; Kopalle, Praveen, K. (2006). The Usefullness of Measuring
Disruptiviness of Innovations Ex Post in Making Ex Ante Predictions, Journal of
Product Innovation Management, Vol. 23, Issue 1, pp. 12-18
196
Greene, Jennifer C.; Caracelli, Valerie J.; Graham, Wendy F. (1989). Toward a
Conceptual Framework for Mixed-Method Evaluation Designs, Educational Evaluation
and Policy Analysis, Vol. 11, Issue 3, pp. 255-274
Grinyer, P.H.; Spender, J.-C., (1979) Recipes, Crises, and Adaptation in Mature
Businesses, International Studies of Management & Organization, Vol. 9, Issue 3, pp.
113-133
Guba, Egon G.; Lincoln, Yvonna S. (1994). Competing Paradigms in Qualitative
Research, Handbook of Qualitative Research, Edited by Denzin, Norman K.; Lincoln
Yvonna S.Sage Publications Ltd.,London, UK, pp. 105-117
Gubbi, Jayavardhana; Buyya, Rajkumar; Marusic, Slaven; Palaniswami, Marimuthu
(2013). Internet of Things (IoT): A Vision, Architectural Elements, and Future Directions,
Future Generation Computer Systems, Vol. 29, Issue 7, pp. 1645-1660
Gupta, A.; Sounder W. (1998). Key Drívers for Reduced Cycle Times, Research
Technology Management, Vol. 41, Issue 4, pp. 38-43
Guszcza, James; Richardson, Bryan (2014). Two Dogmas of Big Data, Understanding
the Power of Analytics for Predicting Human Behavior, Deloitte Review, Issue 15, pp.
163-175
Hacklin, Fredrik; Battistini, Boris, von Krogh, Georg (2013). Strategic Choices in
Converging Industries, MIT Sloan Management Review, Vol. 55, Issue 1, pp. 65-73
Hacklin, Fredrik; Marxt, Christian; Fahmi, Fritz (2010). An Evolutionary Perspective on
Convergence: Including a Stage Model of Inter-Industry Innovation, International
Journal of Technology, Vol. 49 Issue 1-3, pp. 220-249
Hamel, Gary (1996). Strategy as Revolution, Harvard Business Review, Vol. 74, Issue
4, pp. 69-82
Hamel, Gary (1999). Bringing Silicon Valley Inside, Harvard Business Review, Vol. 77,
Issue 5, pp. 70-84
Hammer, Michael (1990). Reengineering Work: Don’t Automate, Obiliterate, Harvard
Business Review, Vol. 68, Issue 4, pp. 104-112
197
Harvey, William S. (2011). Strategies for Conducting Elite Interviews, Quality Research,
Vol. 11, Issue 4, pp. 431-441
Hausmann, Ludwig; Nangia, Ishaan; Netzer, Thomas; Rehm, Werner; Rothkopf,
Maximilian (2015). Pathway to Value Creation, A Perspective on How Transportation
and Logistics Can Increase Their Economic Profit, McKinsey & Company, p. 24
Henderson, Bruce (1970). The Product Portfolio,
https://www.bcgperspectives.com/content/classics/strategy_the_product_portfolio/,
retiereved 6.6.2016
Hendersson, Rebecca, (2006). The Innovator’s Dilemma as a Problem of
Organizational Competence, The Journal of Product Innovation Management, Vol. 23,
Issue 1, pp. 5-11
Hendersson, Rebecca; Clark, Kim B. (1990). Architectural Innovation: The
Reconfiguration of Existing Product Technologies and the Failure of Established Firms,
Administrative Science Quarterly, Vol. 35, Issue 1, pp. 9-30
Hillebrand, Friedhelm (2013). The Creation of Standards for Global Mobile
Communication: GSM and UMTS Standardization 1982 to 2000, IEEE Wireless
Communication, Vol. 20, Issue 5, p. 24 -33
Hofstede, Geert, H.; Hofstede Gert Jan; Minkov, Michael (2010 3rd ed). Cultures and
Organizations: Software of the Mind, McGraw Hill, New York, USA, p. 8-9, 31
Howe, Jeff (2006). The Rise of Crowdsourcing,
http://archive.wired.com/wired/archive/14.06/crowds_pr.html, retrieved 7.7.2015
Huberty, Mark (2015), Awaiting the Second Big Data Revolution: From Digital Noise to
Value Creation, Journal of Industry, Competition and Trade, Vol. 15, Issue 1, pp. 35-47
Huizingh, Eelko K.R.E. (2011). Open Innovation: State of the Art and Future
Perspectives, Technovation, Vol. 31, Issue 1, pp. 2-9
Humphrey, Albert S. (2005). SWOT Analysis for Management Consulting, SRI Alumni
Association Newsletter, December
198
Husig, Stefan; Hipp, Christiane, Dowling, Michael (2005). Analysing Disruptive
Potential: the Case of Wireless Local Area Network and Mobile Communication
Network Companies, R&D Management, Vol. 35, Issue 1, pp. 17-35
Häcki, Remo; Lighton, Julian (2001). The Future of Networked Company, McKinsey
Quarterly, Issue 3, pp. 26-39
INCOSE (2007). Systems Engineering Vision 2020, INCOSE-TP-2004-004-02, Version
2.03, September, p. 16
Ishikawa, Kaouru (1985), What is Total Quality Control? The Japanese Way, Prentice-
Hall, Inc. Englewood Cliffs, N.J., USA, pp. 107-109
Itamar, Arel; Rose Derek C.; Karnowski, Thomas P. (2010). Deep Machine Learning, A
New Frontier in Artificial Intelligence Research, IEEE Computional Intelligence
Magazine, Vol. 5, Issue 4, pp. 13-22
Iyer, Bala; Hendersson, John C. (2010). Preparing to the Future: Understanding the
Seven Capabilities of Cloud Computing, MIS Quarterly Executive, Vol. 9, Issue, 2, pp.
117-131
John, Nicholas (2013). Sharing and Web 2.0: The Emergence of a Keyword, New
Media and Society, Vol. 15, Issue 2, pp. 167-182
Juhanko, Jari (toim.); Jurvansuu, Marko (toim.); Ahlqvist, Toni; Ailisto, Heikki; Alahuhta,
Petteri; Collin, Jari; Halen, Marco; Heikkilä, Tapio; Kortelainen, Helena; Mäntylä, Martti;
Seppälä, Timo; Sallinen, Mikko; Simons, Magnus ja Tuominen, Anu, Suomalainen
(2015), Teollinen internet – haasteesta mahdollisuudeksi: taustoittava kooste, ETLA
Raportit No. 42, https://www.etla.fi/wp-content/uploads/ETLA-Raportit-Reports-42.pdf,
p.9., retrieved 7.7.2015
Kagermann, Henning; Riemensperger, Frank; Hoke, Dirk; Helbig, Johannes;
Stocksmeir, Dirk; Wahlster, Wolfgang; Scheer, August-Wilhelm; Schweer, Dieter
(2014). Smart Service Welt, Recommendations for Implementing the Strategic Initiative
Web-based Services for Businesses,
http://www.acatech.de/fileadmin/user_upload/Baumstruktur_nach_Website/Acatech/roo
t/de/Projekte/Laufende_Projekte/Smart_Service_Welt/BerichtSmartService_engl.pdf,
retrieved 26.5.2016, pp.5-7
199
Kagermann, Henning; Wahlster, Wolfgang; Helbig, Johannes (2013).
Recommendations for Implementing the Strategic Initiative INDUSTRIE 4.0,
http://www.acatech.de/fileadmin/user_upload/Baumstruktur_nach_Website/Acatech/roo
t/de/Material_fuer_Sonderseiten/Industrie_4.0/Final_report__Industrie_4.0_accessible.
pdf, retrieved 12.8.2015, pp. 24-25
Kahneman, Daniel (2003). Maps of Bounded Rationality: Psychology of Behavioral
Economics, American Economic Review, Vol. 93, Issue 5, pp. 1449-1475
Kahneman, Daniel (2011). Thinking Fast and Slow, Farrah, Straus and Giroux, New
York, USA, pp. 129-130, 202-204, 363-374
Kahneman, Daniel; Tvesky, Amos (1979). Prospects Theory: An Analysis of Decision
Under Risk, Econometrica, Vol 47, Issue 2, pp.263-291
Kaplan, Robert S.; Norton, David P. (1992). The Balanced Scorecard – Measures That
Drive Performance, Harvard Business Review, Vol. 70, Issue 1, pp. 71-79
Kellmereit, Danie; Obodovski, Daniel (2013). The Silent Intelligence, DnD Ventures,
San Fransisco, USA, pp. 69-88
Kenney, Martin; Pon, Bryan (2011). Structuring the Smartphone Industry: Is the Mobile
Internet OS Platform the Key, Journal of Industry, Competition and Trade, Vol. 11,
Issue 3, pp. 239-261
Kilby, Jack S. (2000). The Integrated Cicuits Early History, Proceedings of the IEEE,
Vol 88, Issue 1, pp. 109-111
Killing, Peter; Malnight, Thomas; Keys, Tracey (2006). Must Win Battles, How to Win
Them, Again and Again, Pearson Education Inc. Upper Saddle River, NJ, USA , pp.
234-235
Kim W. Chan; Mauborgne, Renee (2005). Blue Ocean Strategy, Harvard Business
School Publishing Corporation, Boston, USA, pp. 16-18, 47-49
Kimble Chris, Bourdon Isabelle (2013). The Link Among Information Technology,
Business Models, and Strategic Breakthroughs: Examples from Amazon, Dell, and
eBay, Global Business and Organizational Excellence, Vol. 33, Issue 1, pp. 58-68
200
Klepper, Steven (1997). Industry Life Cycles, Industrial and Corporate Change, Vol. 6.
Issue 1, pp. 145-181
Kocourek, Paul; Hyde, Paul; Bollinger, Martin; Spiegel, Erik; Treat John (2000).
Beyond Command-and-Control: Managing the Diverse Corporation in Today’s
Turbulent Times, http://www.boozallen.com/media/file/80674.pdf, p. 3
Kohler, William J; Colbert-Taylor, Alex (2015). Current Law and Potential Legal Issues
Pertaining to Automated, Autonomous and Connected Vehicles, Santa Clara Computer
and High – Technology Law Journal, Vol. 31, Issue 1, pp. 99-138
Koivuniemi, Jussi (2013). Liitoa-Kiitoa-Siirtoa, Cargotec Finland Oy, Tampere, Finland
Kostoff, Ronald, N.; Boylan, Robert; Simons, Gene R. (2004). Disruptive Technology
Roadmaps, Technological Forecasting and Social Change, Vol. 71, Issue 1-2, pp. 141-
159
Kuhn, Thomas S. (1996). The Structure of Scientific Revolutions, The University of
Chicago Press, pp. 44, 175
Kushida Kenji E.; Murray, Jonathan; Zysman, John (2015). Cloud Computing: From
Scarcity to Abundance, Journal of Industry, Competition and Trade, Vol.15, Issue 1, pp.
5-19
Kuusi, Osmo (1999). Expertise In the Future Use of Generic Technologies, Epistemic
and Methodological Considerations Concerning Delphi Studies, VATT-Reseach reports
59, Helsinki, pp. 15, 34-36, 77, 128-129
Kuusi, Osmo; Bergman, Timo; Salminen, Hazel (2013). Tulevaisuustietoisuudesta ja
tulevaisuudesta tietämisestä: Finnish Society for Future Studies, Helsinki, Finland, p. 3
Kuusi, Osmo; Kuhl, Kerstin; Steinmuller, Karlheinz (2015). The Futures Map and Its
Quality Criteria, European Journal of Futures Research, Vol 3, Issue 1, pp. 1-14
Laplume, André O; Sonpar Karan; Litz Reginald A. (2008). Stakeholder Theory:
Reviewing a Theory That Moves Us, Journal of Management Vol. 34 Issue 6, pp. 1152-
1189
201
Lee, Gwendolyn K. (2007). The Significance of Network Resources in the Race to
Enter Emerging Product Market: The Convergence of Telephony Communication and
Computer Networking 1989-2001, Strategic Management Journal, Vol 28. Issue 1,
pp.17-37
Lee, Hau L. (2002). Aligning Supply Chain Strategies with Product Uncertainties,
California Management Review, Vol. 44, Issue 3, pp. 105-119
Lee, Sang M.; Olson, David L.; Silvana, Trimi (2010). The Impact of Convergence on
Organizational Innovation, Organizational Dynamics, Vol. 39 Issue 3, pp. 218-225
LeHong, Hung; Fenn, Jackie; Toit Leeb-du, Rand (2014). Hype Cycle for Emerging
Technologies, Gartner, p. 8
Lei, David T. (2000). Industry Evolution and Competence Development: the
Imperatives of Technological Convergence, International Journal of Technology
Management, Vol. 19, Issue. 7/8, pp. 699-738
Levinson, Marc (2006). The Box, How Shipping Container Made the World Smaller and
the World Economy Bigger, Princeton University Press, Princeton, USA, pp. 1-15
Lincoln, Yvonne S. ; Guba, Egon G. (1985). Naturalistic Inquiry. Sage Publications,
Newbury Park, USA, pp. 290-294
Linstone, Harold A (2002). Eight Basic Pitfalls: A checklist, The Delphi Method :
Techniques and Applications, Edited by Linstone, Harold A; Turoff, Murray,
academia.edu, pp. 559-571
Linstone, Harold A; Turoff, Murray (2002). The Delphi Method : Techniques and
Applications, academia.edu, pp. 3-4,
Linton, Jonathan (2002). Forecasting the Market Diffusion of Disruptive and
Discontinuous Innovation, IEEE Transaction on Engineering Management, Vol. 49,
Issue 4, pp. 365-374
Lorange, Peter (2009). Shipping Strategy, Cambridge University Press, New York,
USA, p. 44
202
Low, Murray B.; Abrahamson, Eric (1997). Movements, Bandwagons, and Clones:
Industry Evolution and the Entrepreneurial Process, Journal of Business Venturing Vol.
12, Issue 6, pp.435-457
Lucas Jr. Henry C.; Goh Jie Mein (2009). Disruptive Technology: How Kodak missed
the digital photography revolution, Journal of Strategic Information Systems, Vol. 18,
Issue 1, pp. 46-55
Luostarinen Reijo (1979). Internationalization of the firm: an empirical study of the
internationalization of firms with small and open domestic markets with special
emphasis on lateral rigidity as a behavioral characteristics in strategic decision making,
Acta Academia Oeconomicae Helsingiensis, Series A:30, Helsinki, Finland, p. 182
Machiavelli, Niccoló (2003). Translated by George Bull. The Prince, Penguin Books,
London, UK
Malaska, Pentti (2013). Tulevaisuustietoisuudesta ja tulevaisuudesta tietämisestä:
Tulevaisuus mielenkiinnon kohteena, Edited by Kuusi, Osmo; Bergman, Timo,
Salminen, Hazel, Finnish Society for Future Studies, Helsinki, Finland, p. 19
Malerba, Franco (2007). Innovation and the Dynamics and Evolution of Industries;
Progress and Challenges, International Journal of Industrial Organization, 25, Issue 4,
pp. 675-699
Malerba, Franco; Nelson, Richard; Orsenigo, Luigi; Winter, Sidney (2007). Demand,
Innovation, and the Dynamics of Market Structure: The Role of Experimental Users and
Diverse Preferences, Journal of Evolutionary Economics, Vol 17, Issue 4, pp. 371-399
Malone, Thomas W.; Laubacher, Robert; Dellarocas Chrysanthos (2010). The
Collective Intelligence Genome, MIT Sloan Management Review, Vol. 51, Issue 3, pp.
21-31.
Mannermaa, Mika (1999). Tulevaisuuden hallinta, WSOY, Porvoo, Finland, pp. 147,
153
Manyika, James; Chui, Michael; Bughin, Jacques; Dobbs, Richard; Bisson, Peter;
Mars, Alex (2013). Disruptive Technologies: Advances that will transform life, business
and the global economy, McKinsey Global Institute, May, p. 12, 52
203
Markides, Constantinos (2006). Disruptive Innovations: In Need of Better Theory,
Journal of Product Innovation Management, Vol. 23, Issue 1, pp. 19-25
Martino, Joseph P. (1993). Technological Forecasting for Decision Making, McGraw-
Hill Inc, New York, USA, pp. 38-39
Matthyssens, Paul; Vandenbempt, Koen; Berghman, Liselore (2006). Value Innovation
in Business Markets: Breaking the Industry Recipe, Industrial Marketing Management,
Vol. 35, Issue 6, pp. 751-761
McAfee, Andrew; Brynjolfsson Erik (2012). Big Data: The Management Revolution,
Harvard Business Review, Vol. 90, Issue 10, pp. 60-68
McGahan, Anita M. (2000). How Industries Evolve, Business Strategy Review, Vol. 11,
Issue 3, pp. 1-16
Mell, Peter; Grance, Timothy (2011). The NIST Definition of Cloud Computing, NIST
Special Publication 800-145, September, pp. 2-3
Metcalfe, Bob (2013). Metcalfe’s Law after 40 Years of Ethernet, IEEE Computer,
Volume 46, Issue 12, pp. 26-31
Minzberg, Henry; Ahlstrand, Bruce; Lampel, Joseph (1998). Strategic Safari, The Free
Press, New York, USA, pp.128, 352-359, 369
Minzberg, Henry; Lampel, Joseph (1999). Reflecting on the Strategy Process, MIT
Sloan Management Review, Vol. 40, Issue 3, pp 21-30
Mitchell, Ronald K.; Agle, Bradley R.; Wood, Donna J. (1997). Toward a Theory of
Stakeholder Identification and Salience: Defining the Principle of Who and What Really
Counts, Academy of Management Review, Vol 22, Issue 4, pp. 853-886
Moore, Geoffrey A. (2005). Dealing with Darwin, How Great Companies Innovate at
Every Phase of Their Evolution, Penguin Group, New York, USA, pp. 62-63
Neibecker, Bruno (1984). The Validity of Computer-Controlled Magnitude Scaling to
Measure Emotional Impact of Stimuli, Journal of Marketing Research, Vol. 21, Issue 3,
pp. 325-331
204
Nichols, Keith (2013). White Paper for Smart and Synchronized 3DExperience,
Dassault Systemes, December, http://www.3ds.com/industries/industrial-
equipment/resource-center/white-papers/smart-synchronized/, retrieved 12.8.2015, pp.
2-3
Niiniluoto, Ilkka (2013). Tulevaisuudentutkimus – tiedettä vai taidetta, Miten tutkimme
tulevaisuuksia? Edited by Kuusi, Osmo; Bergman, Timo, Salminen, Hazel, Finnish
Society for Future Studies, Helsinki, Finland, pp.27-28
Nuzzo, Regina (2014). Scientific Method: Statistical Errors, Nature Vol. 506, Issue
7487, pp. 150-152
Näsi, Juha; Aunola, Mauno (2002). Strategisen johtamisen teoria ja käytäntö,
Metalliteollisuuden Keskusliitto, Helsinki, p. 19
Olkkonen, Tauno (1994). Johdatus teollisuustalouden tutkimustyöhön, Helsinki
University of Technology, Report 153, Espoo, Finland, p. 38,39,44
O´Reilly, Tim (2007). What is Web 2.0: Design Patterns and Business Models for the
Next Generation of Software, Communication & Strategies, Q1, Issue 1, pp. 17-37
Osterwalder, Alexander (2004). The Business Model Ontology, A Proposition in a
Design Science Approach, These de l’Ecole des HEC de l’Université de Lausanne, p.
100
Osterwalder, Alexander; Pigneur, Yves (2010). Business Model Generation: A
Handbook for Visionaries, Game Changers, and Challengers, John Wiley & Sons,
Hoboken, USA, pp. 14-17,44
Penrose, Edith (1995 3rd ed.). The Theory of the Growth of the Firm, Oxford University
Press, Oxford, UK, 43-64
Popper, Karl (2002). The Open Society and Its Enemies, Routledge Classics, London,
UK, p. xxxvii
Porter, Michael (1979). How Competitive Forces Shape Strategy, Harvard Business
Review, Vol. 57, Issue 2, pp. 137-145
Porter, Michael (1985), Competitive Advantage – Creating and Sustaining Superior
Performance, The Free Press, New York, USA, pp. 12-17, 37
205
Porter, Michael E.; Heppelman, James, E. (2014). How Smart, Connected Products
Are Transforming Competition, Harvard Business Review, Vol. 92, Issue 11, pp. 64-88
Porter, Michael E.; Heppelman, James, E. (2015). How Smart, Connected Products
Are Transforming Companies, Harvard Business Review, Vol. 93, Issue 10, pp. 96-114
Prahalad, C.K.; Bettis, Richard A. (1986). The Dominant Logic: a New Linkage
Between Diversity and Performance, Strategic Management Journal, Vol. 7, Issue 6,
pp. 485-501
Prahalad, C.K; Hamel Gary (1990). The Core Competence of the Corporation, Harvard
Business Review, Vol. 68, Issue. 3, pp. 79–91.
Praxley, Matt (2007). Honoring the Hard Disk. Information Professional, Vol. 4, Issue 2,
pp. 36-37
Precott, J.E.; Karim S; Hsu S. (2014). The Strategic Process and Competitive
Dynamics of Industry Convergence, Strategic Management Society, 34th Annual
Conference, Madrid
Rao, Bharat; Bojan, Angelov; Nov, Oded (2006). Fusion of Disruptive Technologies:
Lessons from the Skype Case, European Management Journal, Vol. 24, Issue 2-3, pp.
174-188
Razali, Normadiah Mohd; Wah, Yap Bee (2011). Power Comparisons of Shapiro-Wilk,
Kolmogorod-Smirnov, Lillefors and Anderson-Darling Tests, Journal of Statistical
Modeling and Analytics, Vol. 2, Issue 1, pp. 21-33
Ries, Eric (2011). The Lean Startup: How Today's Entrepreneurs Use Continuous
Innovation to Create Radically Successful Businesses, Crown Publishing Group, New
York, USA, pp. 8-9, 111
Rigby, Darrel; Bilodeau, Barbara (2005). The Bain 2005 Management Tool Survey,
Strategy & Leadership, Vol. 33, Issue. 4, pp.4-12
Robertson, David (2006). The Past, Present, and Future of Data Converters and Mixed
Signal ICs: a Universal Model, IEEE Symposium on VLSI Circuits, Digest of Technical
Papers. p. 1-4
206
Robertson, Thomas (1967). The Process of Innovation and the Diffusion of Innovation,
Journal of Marketing, Vol 31, Issue 1, pp. 14-19
Rogers, Everett M. (2003, 5th ed.) The Diffusion of Innovation, Free Press, New York,
USA, p. 281
Roijakkers, Nadine; Zynga, Andy; Bishop, Caroline (2014). Getting Help from
Innomediares, What Can Innovators do to Increase Value in External Knowledge
Searches?, New Frontiers in Open Innovation, Edited by Henry Chesbrough, Wim
Vanhaverbeke and Joel West, Oxford University Press, Oxford, UK, pp. 242-255
Roscoe, J.T. (1975 2nd ed.). Fundamental Research Statistics for the Behavioral
Sciences, Holt, Rinehart & Winston, New York, USA, p.163
Rosenzweig, Phil (2007). The Halo Effect, Free Press, New York, USA, pp.72-75
Rowe, Gene; Wright, George (1999). The Delphi Technique as a Forecasting Tool:
Issues and Analysis, International Journal of Forecasting, Vol. 15, Issue 4, pp. 353-375
Russo, Massimo; Helm, Richard; McGabe, Grant (2014). The Industrial Internet: Six
Critical Question for Equipment Suppliers,
https://www.bcgperspectives.com/content/articles/metals_mining_engineered_products
_infrastructure_industrial_internet/, retrieved 12.8.2015
Sackman, Harold (1974). Delphi Assessment: Expert Opinion, Forecasting, and Group
Process, Rand Corporation, Santa Monica, USA, pp. 69-73
Savitskaya, Irina; Salmi, Pekka; Torkkeli, Marko (2010), Barriers to Open Innovation:
Case China, Journal of Technology Management and Innovation; Vol. 5, Issue 4, pp.
10-21
Sawhney, Mohanbir S. (1998). Leveraged High-Variety Strategies: From Portfolio
Thinking to Platform Thinking, Journal of Academy of Marketing Science, Vol. 26, Issue
1, pp. 54-61
Schein, Edgar H. (1984). Coming to a New Awareness of Organizational Culture, MIT
Sloan Management Review, Vol. 25, Issue 2, pp. 3-16
Schein, Edgar H. (2009 2nd ed.). The Corporate Culture Survival Guide, Jossey-Bass,
San Francisco, USA, p. 6, 27
207
Schmidt, Roy C. (1997). Managing Delphi Surveys Using Nonparametric Statistical
Techniques, Decision Sciences, Vol. 28, Issue 3, pp.763-774
Schumpeter, Joseph A. (2008 3rd ed.). Capitalism, Socialism and Democracy,
HarperCollins Publishers, New York, USA, pp. 81-86
Schumpeter, Joseph A. (2012, 16th ed.), The Theory of Economic Development,
Transaction Publishers, New Brunswick, USA, pp. 57-59
Schwaber, Ken, Beedle, Mike (2002). Agile Software Development with Scrum,
Prentice Hall, Upper Saddle River, USA, pp. 50-56
Schwartz, Shalom H.; Bilsky, Wolfgang (1987). Toward a Universal Psyhological
Strcuture of Human Values, Journal of Personality and Social Psychology, 1987, Vol.
53, Issue 3, pp. 550-562
Schäfer, Daniel (2014). Container Terminal Foresight 2020, Rothschönberg, Germany,
pp. 27,35, 38 / section II and 12,18, 67 / section IV
Shapiro, S.S.; Wilk M.B. (1965). An Analysis of Variance Test for Normality (complete
samples), Biometrika, Vol. 52, Issue 3-4, pp. 591-611
Short, Jeremy C.; Ketchen, David J. Jr.; Palmer, Timothy B. (2002). The Role of
Sampling in Strategic Management Research on Performance: A Two-Study Analysis,
Journal of Management, Vol. 28, Issue, 3, pp. 363-385
Simon, Herbert A. (1979). Rational Decision Making in Business Organizations,
American Economic Review, Vol. 69, Issue 4, pp. 493-513
Simon, Herbert A. (1987). Making Management Decisions: the Role of Intuition and
Emotion, Academy of Management Executive, Vol. 1, Issue 1, pp. 57-64
Simon, Herbert A. (1991). Bounded Rationality and Organizational Learning,
Organization Science, Vol. 2, Issue 1, pp. 125-134
Smircich, Linda; Stubbart, Charles (1985). Strategic Management in an Enacted World,
Academy of Management Review, Vol. 10, Issue 4, pp. 724-736
Smith, Adam (2014), The Wealth of Nations, CreateSpace Independent Publishing,
USA, pp. 3-4
208
Sood, Ashish; Tellis, Gerard J. (2011). Demystifying Disruption: A New Model for
Understanding and Predicting Disruptive Technologies, Marketing Science, Vol. 30,
Issue.2. pp. 339-354
Spence, Michael A. (1981). The Learning Curve and Competition, The Bell Journal of
Economics, Vol. 12, Issue 1, pp. 49-70
Spender J.-C. (1989), Industry Recipes, An Enquiry into Nature and Sources of
Managerial Judgement, http://jcspender.com/uploads/Industry_recipes.pdf, 21.7.2015,
p. 179
Srinivasan, Raji; Haunschild, Pamela; Grewald, Rajdeep (2007). Vicarious Learning in
New Product Introductions in the Early Years of Converging Market, Management
Science, Vol. 53, Issue 1, pp. 16-28
Starbuck, William H. (1982). Congealing Oil: Inventing Ideologies to Justify Acting
Ideologies Out, Journal of Management Studies, Vol. 19 Issue 1, pp. 3-27
Stenbacka, Carolina (2001). Qualitative Research Requires Quality Concepts of Its
Own, Management Decision, Vol. 39, Issue 7, pp. 551-556
Sterman, John D. (2000). System Dynamics: System Thinking and Modelling for a
Complex World. McGraw Hill, New York, USA, pp. 137-141
Stotts, Larry B.; Karp, Sherman; Aein, Joseph M. (2014). The Origins of Miniature
Global Positioning System-Based Navigation Systems. IEEE Signal Processing
Magazine, Vol. 33, Issue 6, pp. 114-117
Stubbart, Charles I. (1989). Managerial Cognition: A Missing Link in Strategic
Management Research, Journal of Management Studies, Vol. 26, Issue 4, pp- 325-347
Suarez, Fernando, F.; Cusamo, Michael A. (2009). The Role of Services in Platform
Markets, Platforms, Markets and Innovation (edited by Gawer, Annabelle), Edward
Elgar, Cheltenham, UK, pp. 77-98
Sun, Tzu (1963). Translated by Samuel B. Griffith. The Art of War, Oxford University
Press, Oxford, UK
Surowiecki, James (2005). The Wisdom of Crowds, Anchor Books, New York, USA, pp.
31-36
209
Taleb, Nassim Nicholas (2010 2nd ed. ). The Black Swan. The Impact of the Highly
Improbable, Random House, New York, USA, pp.349-360
Tannenbaum Robert; Schmidt, Warren H. (1958). How to Choose A Leadership
Pattern, Harvard Business Review, Vol. 36, Issue 2, pp. 95-101
Taylor, Frederick W. (1998). The Priciples of Scientific Management, Dover
Publications, New York, USA, pp.1-11
Teece, David J. (2007). Expliciting Dynamic Capabilites: The Nature and
Microfoundations of (Sustainable) Enterprise Performance, Strategic Management
Journal, Vol. 28, Issue 13, pp. 1319-1350
Tellis, Gerard, J. (2006). Disruptive Technology or Visionary Leadership, Journal of
Product Innovation Management, Vol. 23, Issue 1, pp. 34-38
Tellis, Gerard, J.; Prabhu, Jaideep C..; Chandy Rajesh K. (2009). Radical Innovation
Across Nations: The Preeminence of Corporate Culture, Journal of Marketing, Vol. 73,
Issue 1, pp. 3-23
Tikkanen, Henrikki; Lamberg, Juha-Antti, Parvinen, Petri; Kallunki, Juha-Pekka (2005).
Managerial Coginition, Action and the Business Model of the Firm, Management
Decision, Vol. 43, Issue 6, pp- 789-809
Trigeorgis, Lenos (1993). Real Options and Interactions with Financial Flexibility,
Financial Management, Vol. 22, Issue 3, pp. 202-224
Trott, Paul; Hartmann, Dap (2009). Why “Open Innovation” Is Old Wine in New Bottles,
International Journal of Innovation Management, Vol 13, Issue 4, pp. 715-736
Tushman, Michael L. (1997). Winning Through Innovation, Strategy and Leadership,
Vol. 25, Issue 4, pp. 14-19
UNCTAD (2015). Review of Maritime Transport, United Nations Publications, Geneva,
Switzerland, pp. 22,37
Utterback, James M; Abernathy, William, J. (1975). A Dynamic Model of Process and
Product Innovation, Omega International Journal of Management Science, Vol. 3, Issue
6, pp. 639-656
210
Van der Hejjden, Kees (2002). Scenarios, John Wiley & Sons Ltd, Chichester, England,
pp. 91-93
Vanhaverbeke, Wim; Cloodt, Myriam (2014). Theories of the Firm and Open Innovation,
New Frontiers in Open Innovation, Edited by Henry Chesbrough, Wim Vanhaverbeke
and Joel West, Oxford University Press, Oxford, UK, pp. 256-278
Veryzer, Robert W. (1998). Discontinuos Innovation and the New Product Development
Process, Journal of Product Innovation Management, Vol. 15, Issue 4, pp. 304-321
Viitanen, Mikko; Karvonen, Tapio; Vaiste, Johanna; Hernesniemi, Hannu (2003). Suo-
men Meriklusteri, Teknologiakatsaus 140/2003, Tekes, Helsinki, Finland, p. 165
von Tunzelmann, Nick; Malerba, Franco; Nightingale, Paul; Metcalfe, Stan (2008).
Technological Paradigms: Past, Present and Future. Industrial and Corporate Change,
Vol. 17, Issue 3, pp. 467-484
Walsh, Steven T. (2004). Roadmapping a Disruptive Technology: A case study, The
Emerging Microsystems and Top-Down Nanosystems Industy, Technological
Forecasting and Sosial Change, Vol. 71, Issue 1-2, pp. 161-185
Walshham, Geoff (2006). Doing Interpretive Research, Europen Journal of Information
Systems, Vol 15, Issue 3, pp. 320-330
Waterman, Robert H. Jr.; Peters, Thomas J.; Phillips, Julien R. (1980). Structure Is Not
Organization, Business Horizons, Vol. 23, Issue 3, pp 14-26
Weinhardt, Christof; Anandasivam Wirt, Arun; Blau, Benjamin; Borissov, Nikolay; Meinl,
Thomas; Michalk Wirt, Wibke; Stosser, Jochen (2009). Cloud Computing – A
Classification, Business Models and Research Directions, Business & Information
Systems Engineering, Vol. 1, Issue 5, pp. 391-399
Wenger, Etienne C., Snyder, William M. (2000), Communities of Practise, The
Organizational Frontier, Harvard Business Review, Vol. 78 Issue 1, pp. 139-145
Wessel, Maxwell; Christensen, Clayton M. (2012). Surviving Disruption, Harvard
Business Review, Vol 90, Issue 12, pp. 56-64
211
West, Joel; Gallagher, Scott (2006), Challenges of Open Innovation: The Paradox of
Firm Investment in Open-source Software, R&D Management, Vol. 36, Issue 3, pp.
319-331
Westerman, George; Bonnet, Didier; McAfee, Andrew (2014). Leading Digital, Turning
Technology into Business Transformation, Harvard Business Review Press, Boston,
USA, pp. 39-42
Westerman, George; Tannou, Maöl; Bonnet, Didier; Ferraris, Patrick; McAfee, Andrew
(2012). The Digital Advantage, How Digital Leaders Outperform Their Peers in Every
Industry,https://www.capgemini.com/resource-file-access/resource/pdf/The_Digital_
Advantage__How_Digital_Leaders_Outperform_their_Peers_in_Every_Industry.pdf,
p.9, Retrieved 6.8.2015
Westley, Frances; Minzberg, Henry (1989). Visionary Leadership and Strategic Man-
agement, Strategic Management Journal, Vol. 10, Supplement Issue, pp. 17-32
de Witt, Rob; Meyer, Ron (2005). Strategy Synthesis, Thomson Learning, London, UK
World Economic Forum (2015). Industrial Internet of Things: Unleashing the Potential
of Connected Products and Services,
http://www3.weforum.org/docs/WEFUSA_IndustrialInternet_Report2015.pdf, retrieved
6.6.2016 pp. 8-9
Yin, Robert K. (2014). Case Study Research: Design and Methods, SAGE Ltd. London,
UK, p.10
Zook, Chris; Allen, James (2012). Repeatability, Harvard Business School Publishing,
Boston, USA, pp. 18-23
Zysman, John; Feldman, Stuart; Kushida Kenji E.; Murran Jonathan; Nielsen Niels
Christian (2013). Services with Everything: The ICT-Enabled Digital Transformation of
Services. The Third Globalization, edited by Breznitz, Dan and Zysman, John, Oxford
Universtiy Press, Oxford UK, pp. 99-129
212
Appendix 1. Mapping strategic schools related to strategy formation (Minzberg et al. 1998)
Rational Natural Internal process
proposed as
Comprehensible, controllable
Unpredictable, confusing
External world seen as
Entrepreneurial Design Positioning
Planning
Environmental Cognitive
Learning
evolutionary and chaos theory
Power (micro)
Cultural
Configuration Power (macro)
213
Appendix 2. Delphi round two interview framework
Research question
It is likely that digitalization is a major systemic enabler for future competitive advantages of the machine-building in-dustry, and in some phase of the value creation, it is or will be disruptive within the next 10 years. Which drivers enabled by digitalization accelerate or inhibit disruption in a machine-building industry that serves global container handling? Do the enablers, actors and users have different views on the drivers or sub-drivers that either accelerate or inhibit the disruptive development? How are disruptive drivers related to each other in a ma-chine-building industry that serves global container han-dling?
214
Products
Services
Object of a potential impact
Operations
Business models
Business model change can also be integrated into prod-ucts, services or operations and there is ongoing diffu-sion of all of the elements like service of everything
215
Continuous improve-ment
Quantum leap
Disruption
Staying where you are
- This is housekeeping
Change of positions
- Innovation gets you here
Change of rules
- Transformation gets you here
Nature of the potential impact
216
Research design
Industry Evolution
Economics Strategy formation
Top Management
Digitalization 3 Cases
Survey
Delphi I Inter-view
Delphi II Inter-view
System model
217
Panel composition
4 4 4
4 4 4
4 4 4
Enabler Actor User
Decision maker
Synthesizer
Scientist
218
Interview process
Intro Technology
rating
Economy
Rating
Factor rating with
Argumentation
Factor rating with argumentation
In order of importance (1-6) from your perspective and view what is their potential to cause disruption
In order of importance (1-7) from your per-spective and view what is their potential to cause disruption
What accelerates?
What slows down?
What accelerates?
What slows down?
219
PEST
Disruption?
220
IoT
+ -
Sub-driver analysis
221
Appendix 3. Survey form
222
Appendix 4. Participants in the Delphi round one and/or two
Dr. Erkki Ahola, Vice President, VTT Ltd (former Vice President, R&D, Sandvik Mining and Construction); (A)
Mr. Kimmo Alkio, President and CEO, Tieto Corporation; (E)
Mr. Timo Antila, former Managing Director, Packaging Industry of M-real (Metsä-Group); (U)
Mr. Andrew Barrons, Senior Vice President and Chief Marketing Officer, Navis LLC.;(E)
Mr. Philippe Bartissol, Vice President, Industrial Equipment, Dassault Systems SE; (E)
Mr. Alex Duca, Director, Head of Design and Automation, APM Terminals; (U)
Mr. Rob van Eijndhoven, former Managing Director, ECT-DELTA Container Division; (U)
Professor Kalevi Ekman, Aalto University; (A)
Mr. Sauli Eloranta, Senior Vice President, Technology Management & Innovation, Rolls Royce Marine; (A)
Chancellor Emeritus Jarl-Thore Eriksson, Åbo Akademi; (E)
Professor Stuart Evans, Carnegie Mellon University; (E)
Mr. Greg Franklin, Co-founder, Intellect Partners; (E)
Mr. Max Furmanov, Managing Director, Accenture Technology; (E)
Mr. Christer Granskog, Chairman of the Board, Patria Corporation; (A)
Mrs. Jena Holtberg-Benge, Director, John Deere WorkSight™, Deere & Co.; (A)
Mr. Robert Inchausti, CTO, XVELA; (E)
Mr. Juhani Jantunen, Technical Director, Dassault Systems; (E)
Mr. Jouko Jokinen, Editor-In-Chief, Aamulehti; (U)
Mr. Pete Karns, Vice President, Internet of Things, IBM: (E)
Dr. Antti Kaunonen, President, Kalmar; (A)
Professor Martin Kenney, UC Davis; (U)
Mr. Frank Kho, former Head of HNN Container Division, PSA; (U)
Prof. Dr. Jürgen Kluge; (A)
Mr. Glen Koskela, CTO, Fujitsu Nordic; (E)
Mr. Mikko Kuitunen, Founder, Vincit Oy; (E)
Dr. Harri Kulmala, CEO, Fimecc Oy; (A)
Mr. Tommi Lehtonen, Senior Vice President, Crushing and Screening Business Line, Metso Corporation; (A)
223
Mr. Edmond Leung, CIO & Head of Operations Development and Technology, Hutchison Port Holdings; (U)
Dr. Raymond E. Levitt, Kumagai Professor of Engineering, Stanford University; (U)
Mr. Perttu Louhiluoto, President, Services, Metso Corporation; (A)
Mr. Harri Luuppala, Co-founder, Open Invest (former Senior Vice President, Tieto Cor-poration); (U)
Mr. Anthony Marshall, Research Director and Strategy Leader, IBM Institute for Busi-ness Value; (E)
Professor Jeff Martin, Chung-Ang University; (U)
Mr. Brendan McDonnell, General Manager, Technology & Engineering, Patrick; (U)
Professor Lasse Mitronen, Aalto University; (U)
Professor Martti Mäntylä, Aalto University; (E)
Professor Jorma Mäntynen, Tampere University of Technology; (U)
Mr. Daniel Navarro, Google for Entrepreneurs; (E)
Vuorineuvos Dr. Kari Neilimo, Director of EMBA program of University of Tampere and Tampere University of Technology (former CEO and Chairman of S Group); (U)
Dr. Markus Nordberg, Head of IdeaSquare, CERN; (A)
Mr. Oh Bee Lock, Head of Group Technology (former COO, Singapore Terminals), PSA; (U)
Dr. Tero Ojanperä, Co-founder, Vision+ (former CTO, Nokia); (E)
Mr. Jari Rautjärvi, Managing Director, Valtra Inc.; (A)
Dr. Petri Rouvinen, CEO, Etlatieto Oy; (A)
Mr. Patrick Scaglia, Co-founder CITRIS Foundry accelerator (former CTO, Hewlett-Packard); (E)
Professor Emeritus Reijo Tuokko, Tampere University of Technology; (A)
Mr. Jan Waas, Director, Technology and Engineering, ECT; (U)
Mr. Bill Walsh, President and CEO, Navis; (E)
Professor Matti Vilkko, Tampere University of Technology; (A)
Professor Kim Wikström, Åbo Akademi; (U)
Professor Emeritus John Zysman, UC Berkeley; (E)
Four panelists opted to stay anonymous: 1 Research director from a European re-search institute (E), 1 Professor from a German university (A), 1 Executive from a German machine builder (A) and 1 Executive from Nordic based process industry (U).
224
Appendix 5. Results from Shapiro-Wilk test
Whole population
Impact Business
Products Service Operations models
p-value < 0,0001 < 0,0001 < 0,0001 < 0,0001
alpha 0,05 0,05 0,05 0,05
Normal No No No No
Technologies 3D Big Data Cloud IoT II MBSE OI Robotics
p-value < 0,0001 < 0,0001 < 0,0001 < 0,0001 0,000 < 0,0001 < 0,0001 < 0,0001
alpha 0,05 0,05 0,05 0,05 0,05 0,05 0,05 0,05
Normal No No No No No No No No
By value chain
Impact Business
Enabler Products Service Operations models
p-value 0,000 < 0,0001 < 0,0001 0,000
alpha 0,05 0,05 0,05 0,05
Normal No No No No
Business
Actor Products Service Operations models
p-value < 0,0001 < 0,0001 < 0,0001 < 0,0001
alpha 0,05 0,05 0,05 0,05
Normal No No No No
Business
User Products Service Operations models
p-value 0,006 0,006 0,029 0,000
alpha 0,05 0,05 0,05 0,05
Normal No No No No
Technologies Enabler 3D Big Data Cloud IoT II MBSE OI Robotics
p-value < 0,0001 < 0,0001 < 0,0001 0,001 0,000 < 0,0001 < 0,0001 0,000
alpha 0,05 0,05 0,05 0,05 0,05 0,05 0,05 0,05
Normal No No No No No No No No
Actor 3D Big Data Cloud IoT II MBSE OI Robotics
p-value < 0,0001 < 0,0001 < 0,0001 < 0,0001 < 0,0001 < 0,0001 < 0,0001 < 0,0001
alpha 0,05 0,05 0,05 0,05 0,05 0,05 0,05 0,05
Normal No No No No No No No No
User 3D Big Data Cloud IoT II MBSE OI Robotics
p-value 0,001 0,005 0,001 0,050 0,010 0,010 < 0,0001 0,000
alpha 0,05 0,05 0,05 0,05 0,05 0,05 0,05 0,05
Normal No No No No No No No No
Test interpretation: H0: The variable from which the sample was extracted follows a Normal distribution.
Ha: The variable from which the sample was extracted does not follow a Normal distribution.
As the computed p-value is lower than the significance level alpha=0,05, one should reject the
null hypothesis H0, and accept the alternative hypothesis Ha.
The risk to reject the null hypothesis H0 while it is true is lower than 0,01%.
225
Appendix 6. Interaction effect test
Goo
dnes
s o
f fi
t b
y ad
just
ed R
²
Prod
ucts
Se
rvic
es
Ope
rati
ons
Bus
ines
s m
ode
ls
3D
Big
Dat
a Cl
oud
Io
T II
MB
SE
OI
Ro
boti
cs
Adj
uste
d R
² 0,
104
0,10
7 0,
069
0,07
3 0,
215
0,09
5 0,
072
0,10
8 0,
134
0,12
9 0,
083
0,07
6 p-
valu
e 0,
002
0,02
3 0,
052
0,00
8 <
0,00
01
0,05
9 0,
043
0,00
1 0,
002
0,00
3 0,
142
0,00
6
Type
III S
S
Prod
ucts
Se
rvic
es
Ope
rati
ons
Bus
ines
s m
ode
ls
3D
Big
Dat
a Cl
oud
Io
T II
MB
SE
OI
Ro
boti
cs
VC*
Rol
e 0,
779
0,93
1 0,
966
0,
674
0,99
3
0,
988
V
C*In
vest
men
t 0,
979
0,51
0 0,
245
0,35
5
0,72
0 0,
669
0,95
3
0,
959
0,74
9 R
ole*
Inve
stm
ent
0,83
2
0,79
0 0,
888
0,
661
0,90
0
VC*
Rol
e*In
vest
men
t
0,06
5
0,03
7
0,
002
0,02
3 0,
053
Expl
anat
ions
:
Type
III S
S va
lues
are
use
d to
vis
ualiz
e th
e in
fluen
ce t
hat
rem
ovin
g an
exp
lana
tory
var
iabl
e ha
s on
the
fit
ting
of t
he m
odel
, all
othe
r va
riab
les
bein
g re
tain
ed, a
s re
gard
s th
e su
m o
f the
squ
ares
of t
he e
rror
s (S
SE),
the
mea
n sq
uare
d er
ror
(MSE
), F
ishe
r's
F, o
r th
e pr
obab
ility
ass
ocia
ted
wit
h
Fish
er's
F. T
he lo
wer
the
pro
babi
lity,
the
larg
er t
he c
ontr
ibut
ion
of t
he v
aria
ble
to t
he m
odel
, all
the
othe
r va
riab
les
alre
ady
bein
g in
the
mod
el.
226
Appendix 7. Disruption impact by whole population, unbalanced data
227
Appendix 8. Disruption impact by value chain, balanced data
228
Appendix 9.Disruptiveness of individual technologies or concepts by value chain, balanced data
229
230
Appendix 10. Correlation matrix (Kendall) with whole population
Balanced data
Impact Technologies
Variables Products Service Operations Business models 3D Big Data Cloud IoT II MBSE OI Robotics
Imp
act
Products x - - - - - - - - - - -
Service 0,39 x - - - - - - - - - -
Operations 0,13 0,19 x - - - - - - - - -
Business models 0,26 0,28 0,22 x - - - - - - - -
Tech
no
logi
es
3D 0,11 0,04 0,10 0,04 x - - - - - - -
Big Data 0,23 0,33 0,26 0,27 -0,04 x - - - - - -
Cloud 0,24 0,20 0,16 0,22 -0,12 0,26 x - - - - -
IoT 0,22 0,25 0,18 0,21 -0,08 0,32 0,29 x - - - -
II 0,18 0,23 0,11 0,27 -0,08 0,26 0,21 0,44 x - - -
MBSE 0,00 0,04 0,07 -0,04 0,14 0,06 -0,10 0,03 0,13 x - -
OI 0,24 0,25 0,07 0,15 0,11 0,18 0,21 0,11 0,13 0,03 x -
Robotics 0,13 0,05 0,13 0,04 0,18 0,09 0,02 0,05 -0,05 0,05 0,18 x
Values in bold are different from 0 with a significance level alpha=0,05
Code Coefficient
0,2 - 0.29 Low correlation
0,3 - 0,39 Medium correlation
> 0,4 High correlation
Unbalanced data
Impact Technologies
Variables Products Service Operations Business models 3D Big Data Cloud IoT II MBSE OI Robotics
Imp
act Products x - - - - - - - - - - -
Service 0,28 x - - - - - - - - - -
Operations 0,13 0,25 x - - - - - - - - -
Business models 0,19 0,26 0,20 x - - - - - - - -
Tech
no
logi
es
3D 0,11 0,07 0,16 0,04 x - - - - - - -
Big Data 0,20 0,30 0,15 0,17 -0,06 x - - - - - -
Cloud 0,15 0,18 0,10 0,13 -0,09 0,30 x - - - - -
IoT 0,16 0,23 0,12 0,19 -0,01 0,25 0,27 x - - - -
II 0,19 0,24 0,16 0,29 0,02 0,29 0,19 0,48 x - - -
MBSE 0,12 0,08 0,09 0,00 0,11 0,09 0,04 0,06 0,10 x - -
OI 0,22 0,21 0,12 0,19 0,15 0,16 0,11 0,09 0,08 0,14 x -
Robotics 0,17 0,12 0,08 0,08 0,20 0,05 0,07 0,10 0,03 0,12 0,20 x
Values in bold are different from 0 with a significance level alpha=0,05
Code Coefficient
0,2 - 0.29 Low correlation
0,3 - 0,39 Medium correlation
> 0,4 High correlation
231
Appendix 11. Correlation matrix (Kendall) by value chain with balanced data
Enablers
Areas Technologies
Variables Products Service Operations Business models 3D Big Data Cloud IoT II MBSE OI Robotics
Are
as
Products x - - - - - - - - - - -
Service 0,40 x - - - - - - - - - -
Operations 0,18 0,23 x - - - - - - - - -
Business models 0,23 0,21 0,20 x - - - - - - - -
Tech
no
logi
es
3D 0,03 0,05 0,08 -0,04 x - - - - - - -
Big Data 0,36 0,32 0,20 0,26 -0,02 x - - - - - -
Cloud 0,12 0,15 0,24 0,12 -0,11 0,19 x - - - - -
IoT 0,20 0,17 0,20 0,18 -0,02 0,34 0,31 x - - - -
II 0,24 0,21 0,07 0,35 -0,02 0,35 0,27 0,50 x - - -
MBSE 0,09 0,05 -0,03 -0,14 0,15 -0,03 -0,01 0,02 0,02 x - -
OI 0,15 0,07 0,03 0,17 0,11 0,15 0,11 0,17 0,26 0,07 x -
Robotics 0,06 -0,02 -0,15 0,09 0,22 0,08 -0,08 0,06 0,14 0,11 0,17 x
Values in bold are different from 0 with a significance level alpha=0,05
Actors
Areas Technologies
Variables Products Service Operations Business models 3D Big Data Cloud IoT II MBSE OI Robotics
Are
as
Products x - - - - - - - - - - -
Service 0,10 x - - - - - - - - - -
Operations 0,15 0,25 x - - - - - - - - -
Business models 0,03 0,17 0,20 x - - - - - - - -
Tech
no
logi
es
3D 0,18 0,05 0,22 -0,01 x - - - - - - -
Big Data 0,07 0,26 0,13 0,04 -0,12 x - - - - - -
Cloud -0,07 0,17 0,05 0,17 -0,07 0,27 x - - - - -
IoT 0,03 0,16 0,05 0,21 0,09 0,18 0,24 x - - - -
II 0,16 0,28 0,16 0,15 0,06 0,36 0,10 0,52 x - - -
MBSE 0,21 0,27 0,16 0,05 0,08 0,27 0,07 -0,03 0,07 x - -
OI 0,16 0,24 0,24 0,25 0,25 0,10 0,03 0,09 0,01 0,23 x -
Robotics 0,08 0,15 0,07 -0,06 0,08 0,02 0,14 0,14 -0,04 0,11 0,26 x
Values in bold are different from 0 with a significance level alpha=0,05
Users
Areas Technologies
Variables Products Service Operations Business models 3D Big Data Cloud IoT II MBSE OI Robotics
Are
as
Products x - - - - - - - - - - -
Service 0,39 x - - - - - - - - - -
Operations 0,24 0,28 x - - - - - - - - -
Business models 0,34 0,37 0,28 x - - - - - - - -
Tech
no
logi
es
3D 0,14 0,17 0,23 0,11 x - - - - - - -
Big Data 0,28 0,30 0,17 0,23 0,06 x - - - - - -
Cloud 0,25 0,21 0,02 0,16 -0,11 0,31 x - - - - -
IoT 0,28 0,35 0,17 0,39 0,00 0,28 0,24 x - - - -
II 0,18 0,22 0,21 0,40 0,13 0,15 0,19 0,54 x - - -
MBSE 0,03 -0,05 0,16 0,00 0,24 0,06 -0,13 0,13 0,17 x - -
OI 0,28 0,28 0,22 0,19 0,18 0,25 0,12 0,19 0,12 0,19 x -
Robotics 0,25 0,15 0,23 0,06 0,12 0,02 0,13 0,06 -0,03 0,17 0,17 x
Values in bold are different from 0 with a significance level alpha=0,05
232
Appendix 12. Variance dependency by ANOVA
Products
Enabler Actor User
3D
Enabler Actor User
Big data
Enabler Actor User
Cloud
Enabler Actor User
IoT
Enabler Actor User
Robotics
Enabler Actor User
Share the same opinion
Differentiates by being the highest value
Differentiates by being the lowest value
Differentiates by being the middle value
Stands somewhere
between and does not
differentiate clearly from the others
Impact
Technologies
Explanation
0,061
0,009
0,006
0,002
0,466
0,259
0,098
0,007
0,186
0,007
0,024
0,894
0,513
0,001
0,008
0,214
0,011
0,130
ISBN 978-952-15-3855-1ISSN 1459-2045
Tampereen teknillinen yliopisto PL 52733101 Tampere
Tampere University of TechnologyP.O.B. 527FI-33101 Tampere, Finland
TTY Julkaisu 1436 • TUT P
ublication 1436 Matti S
omm
arberg